how does statistics help in research

Effective Use of Statistics in Research – Methods and Tools for Data Analysis

Remember that impending feeling you get when you are asked to analyze your data! Now that you have all the required raw data, you need to statistically prove your hypothesis. Representing your numerical data as part of statistics in research will also help in breaking the stereotype of being a biology student who can’t do math.

Statistical methods are essential for scientific research. In fact, statistical methods dominate the scientific research as they include planning, designing, collecting data, analyzing, drawing meaningful interpretation and reporting of research findings. Furthermore, the results acquired from research project are meaningless raw data unless analyzed with statistical tools. Therefore, determining statistics in research is of utmost necessity to justify research findings. In this article, we will discuss how using statistical methods for biology could help draw meaningful conclusion to analyze biological studies.

Table of Contents

Role of Statistics in Biological Research

Statistics is a branch of science that deals with collection, organization and analysis of data from the sample to the whole population. Moreover, it aids in designing a study more meticulously and also give a logical reasoning in concluding the hypothesis. Furthermore, biology study focuses on study of living organisms and their complex living pathways, which are very dynamic and cannot be explained with logical reasoning. However, statistics is more complex a field of study that defines and explains study patterns based on the sample sizes used. To be precise, statistics provides a trend in the conducted study.

Biological researchers often disregard the use of statistics in their research planning, and mainly use statistical tools at the end of their experiment. Therefore, giving rise to a complicated set of results which are not easily analyzed from statistical tools in research. Statistics in research can help a researcher approach the study in a stepwise manner, wherein the statistical analysis in research follows –

1. Establishing a Sample Size

Usually, a biological experiment starts with choosing samples and selecting the right number of repetitive experiments. Statistics in research deals with basics in statistics that provides statistical randomness and law of using large samples. Statistics teaches how choosing a sample size from a random large pool of sample helps extrapolate statistical findings and reduce experimental bias and errors.

2. Testing of Hypothesis

When conducting a statistical study with large sample pool, biological researchers must make sure that a conclusion is statistically significant. To achieve this, a researcher must create a hypothesis before examining the distribution of data. Furthermore, statistics in research helps interpret the data clustered near the mean of distributed data or spread across the distribution. These trends help analyze the sample and signify the hypothesis.

3. Data Interpretation Through Analysis

When dealing with large data, statistics in research assist in data analysis. This helps researchers to draw an effective conclusion from their experiment and observations. Concluding the study manually or from visual observation may give erroneous results; therefore, thorough statistical analysis will take into consideration all the other statistical measures and variance in the sample to provide a detailed interpretation of the data. Therefore, researchers produce a detailed and important data to support the conclusion.

Types of Statistical Research Methods That Aid in Data Analysis

Statistical analysis is the process of analyzing samples of data into patterns or trends that help researchers anticipate situations and make appropriate research conclusions. Based on the type of data, statistical analyses are of the following type:

1. Descriptive Analysis

The descriptive statistical analysis allows organizing and summarizing the large data into graphs and tables . Descriptive analysis involves various processes such as tabulation, measure of central tendency, measure of dispersion or variance, skewness measurements etc.

2. Inferential Analysis

The inferential statistical analysis allows to extrapolate the data acquired from a small sample size to the complete population. This analysis helps draw conclusions and make decisions about the whole population on the basis of sample data. It is a highly recommended statistical method for research projects that work with smaller sample size and meaning to extrapolate conclusion for large population.

3. Predictive Analysis

Predictive analysis is used to make a prediction of future events. This analysis is approached by marketing companies, insurance organizations, online service providers, data-driven marketing, and financial corporations.

4. Prescriptive Analysis

Prescriptive analysis examines data to find out what can be done next. It is widely used in business analysis for finding out the best possible outcome for a situation. It is nearly related to descriptive and predictive analysis. However, prescriptive analysis deals with giving appropriate suggestions among the available preferences.

5. Exploratory Data Analysis

EDA is generally the first step of the data analysis process that is conducted before performing any other statistical analysis technique. It completely focuses on analyzing patterns in the data to recognize potential relationships. EDA is used to discover unknown associations within data, inspect missing data from collected data and obtain maximum insights.

6. Causal Analysis

Causal analysis assists in understanding and determining the reasons behind “why” things happen in a certain way, as they appear. This analysis helps identify root cause of failures or simply find the basic reason why something could happen. For example, causal analysis is used to understand what will happen to the provided variable if another variable changes.

7. Mechanistic Analysis

This is a least common type of statistical analysis. The mechanistic analysis is used in the process of big data analytics and biological science. It uses the concept of understanding individual changes in variables that cause changes in other variables correspondingly while excluding external influences.

Important Statistical Tools In Research

Researchers in the biological field find statistical analysis in research as the scariest aspect of completing research. However, statistical tools in research can help researchers understand what to do with data and how to interpret the results, making this process as easy as possible.

1. Statistical Package for Social Science (SPSS)

It is a widely used software package for human behavior research. SPSS can compile descriptive statistics, as well as graphical depictions of result. Moreover, it includes the option to create scripts that automate analysis or carry out more advanced statistical processing.

2. R Foundation for Statistical Computing

This software package is used among human behavior research and other fields. R is a powerful tool and has a steep learning curve. However, it requires a certain level of coding. Furthermore, it comes with an active community that is engaged in building and enhancing the software and the associated plugins.

3. MATLAB (The Mathworks)

It is an analytical platform and a programming language. Researchers and engineers use this software and create their own code and help answer their research question. While MatLab can be a difficult tool to use for novices, it offers flexibility in terms of what the researcher needs.

4. Microsoft Excel

Not the best solution for statistical analysis in research, but MS Excel offers wide variety of tools for data visualization and simple statistics. It is easy to generate summary and customizable graphs and figures. MS Excel is the most accessible option for those wanting to start with statistics.

5. Statistical Analysis Software (SAS)

It is a statistical platform used in business, healthcare, and human behavior research alike. It can carry out advanced analyzes and produce publication-worthy figures, tables and charts .

6. GraphPad Prism

It is a premium software that is primarily used among biology researchers. But, it offers a range of variety to be used in various other fields. Similar to SPSS, GraphPad gives scripting option to automate analyses to carry out complex statistical calculations.

This software offers basic as well as advanced statistical tools for data analysis. However, similar to GraphPad and SPSS, minitab needs command over coding and can offer automated analyses.

Use of Statistical Tools In Research and Data Analysis

Statistical tools manage the large data. Many biological studies use large data to analyze the trends and patterns in studies. Therefore, using statistical tools becomes essential, as they manage the large data sets, making data processing more convenient.

Following these steps will help biological researchers to showcase the statistics in research in detail, and develop accurate hypothesis and use correct tools for it.

There are a range of statistical tools in research which can help researchers manage their research data and improve the outcome of their research by better interpretation of data. You could use statistics in research by understanding the research question, knowledge of statistics and your personal experience in coding.

Have you faced challenges while using statistics in research? How did you manage it? Did you use any of the statistical tools to help you with your research data? Do write to us or comment below!

Frequently Asked Questions

Statistics in research can help a researcher approach the study in a stepwise manner: 1. Establishing a sample size 2. Testing of hypothesis 3. Data interpretation through analysis

Statistical methods are essential for scientific research. In fact, statistical methods dominate the scientific research as they include planning, designing, collecting data, analyzing, drawing meaningful interpretation and reporting of research findings. Furthermore, the results acquired from research project are meaningless raw data unless analyzed with statistical tools. Therefore, determining statistics in research is of utmost necessity to justify research findings.

Statistical tools in research can help researchers understand what to do with data and how to interpret the results, making this process as easy as possible. They can manage large data sets, making data processing more convenient. A great number of tools are available to carry out statistical analysis of data like SPSS, SAS (Statistical Analysis Software), and Minitab.

nice article to read

Holistic but delineating. A very good read.

Rate this article Cancel Reply

Your email address will not be published.

Enago Academy's Most Popular Articles

• Promoting Research
• Thought Leadership
• Trending Now

How Enago Academy Contributes to Sustainable Development Goals (SDGs) Through Empowering Researchers

The United Nations Sustainable Development Goals (SDGs) are a universal call to action to end…

• Reporting Research

Research Interviews: An effective and insightful way of data collection

Research interviews play a pivotal role in collecting data for various academic, scientific, and professional…

Planning Your Data Collection: Designing methods for effective research

Planning your research is very important to obtain desirable results. In research, the relevance of…

• Language & Grammar

Best Plagiarism Checker Tool for Researchers — Top 4 to choose from!

While common writing issues like language enhancement, punctuation errors, grammatical errors, etc. can be dealt…

• Industry News
• Publishing News

2022 in a Nutshell — Reminiscing the year when opportunities were seized and feats were achieved!

It’s beginning to look a lot like success! Some of the greatest opportunities to research…

2022 in a Nutshell — Reminiscing the year when opportunities were seized and feats…

Sign-up to read more

Subscribe for free to get unrestricted access to all our resources on research writing and academic publishing including:

• 2000+ blog articles
• 50+ Webinars
• 10+ Expert podcasts
• 50+ Infographics
• 10+ Checklists
• Research Guides

We hate spam too. We promise to protect your privacy and never spam you.

I am looking for Editing/ Proofreading services for my manuscript Tentative date of next journal submission:

As a researcher, what do you consider most when choosing an image manipulation detector?

• Spartanburg Community College Library
• SCC Research Guides

Finding Statistics

• Why are Statistics Important?

Statistics are important because they help people make informed decisions. Governments, organizations, and businesses all collect statistics to help them track progress, measure performance, analyze problems, and prioritize. For example, the U.S. Census Bureau collects information from people about where they live and their age. This information can help cities decide where they should build a new hospital if they find that there is a high elderly population in an area or a new school, if they find there are many families with young children.

On a personal level, statistics can be a great way to enhance your argument in a research paper or presentation. They show that there is evidence to back up your claim and can add credibility to your work. Statistics often create an emotional response in your audience. Think about how you feel when someone can back up their argument with statistics? Don't the statistics make you feel more strongly to the argument?

The below video by Ms. Emma Stevenson will help explain how statistics can help you in a research paper or project:

Misleading Statistics

Statistics are an excellent way to enhance an argument and persuade your audience; however, there are some considerations to keep in mind. Statistics can be misleading, because they are often taken out of context. Sometimes, important information is left out about how the statistic was collected in order to make it seem more dramatic, proving big ideas or generalizations that it wouldn't if the rest of the information was included. 

For example, let's say you found a statistic that said 5 out of 5 dentists recommend a certain brand of toothpaste. That sounds like this is a great brand of toothpaste that everyone should use. However, what if you found out that the dentists were all asked if they would recommend that brand of toothpaste or not brushing your teeth at all? Of course all of the dentists are going to pick the brand of toothpaste. This makes the 5 out of 5 recommendation basically meaningless. You might assume when you see this statistic that dentists were ranking this toothpaste brand over other toothpaste brands, instead of against not brushing your teeth at all; this makes the statistic misleading.

Another way statistics can be misleading is in the sample size that the data was collected in. For example, let's say you found a statistic that says 4 out of 5 women prefer wearing high heels over flats to work. However, when you start looking closer at the source the statistic came from, you find that this statistic came from someone asking 5 women they work with in a corporate law firm if they liked wearing heels or flats to work. This is a problem for several reasons.

First, the information was collected from a very small sample size (5 women who all work at the same place). These 5 women cannot represent all women and their opinions on high heels. Second, this sample is very biased, because all of the women work in the same corporate law firm. These women's opinions are not going to reflect all women's opinions, regardless of the number of women sampled, because the women are too similar to one another. If all women in all industries were surveyed for this question, the statistic would look very different. Because of this, it's always important to know the context of any statistic before you use it in your argument. Similarly, you want to be wary of statistics you find that don't have context or can't be tracked back to an original source.

Just like evaluating the credibility of your sources , you will want to do the same for when you want to use statistics in your research. Ask yourself the following questions:

• Can you find the original source that this statistic was published in? This will help you understand the context of the statistics.
• Who published the original source and where was it published?
• Who collected the information for the statistics? Do they have any kind of agenda/stake in the statistics?
• When was the information collected? Could it be out of date?
• How big was the sample size/how much data was collected? What were the demographics of the sample size? This will help you figure out if the statistics are representative of a certain group or area. 

Here is an article that goes deeper into how statistics can be misleading and ways to determine whether your statistics are misleading or not.

• << Previous: Home
• Next: Find Articles (Databases) >>
• Find Articles (Databases)
• Find Websites

Questions? Ask a Librarian

• Last Updated: May 8, 2024 9:31 AM
• URL: https://libguides.sccsc.edu/finding-statistics

Giles Campus | 864.592.4764 | Toll Free 866.542.2779 | Contact Us

Copyright © 2024 Spartanburg Community College. All rights reserved.

Info for Library Staff | Guide Search

Return to SCC Website

User Preferences

Content preview.

Arcu felis bibendum ut tristique et egestas quis:

• Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris
• Duis aute irure dolor in reprehenderit in voluptate
• Excepteur sint occaecat cupidatat non proident

Keyboard Shortcuts

1.1 - what is the role of statistics in clinical research, what is the role of statistics in clinical research section  .

Clinical research involves investigating proposed medical treatments, assessing the relative benefits of competing therapies, and establishing optimal treatment combinations . Clinical research attempts to answer questions such as “should a man with prostate cancer undergo radical prostatectomy or radiation or watchfully wait?” and “is the incidence of serious adverse effects among patients receiving a new pain-relieving therapy greater than the incidence of serious adverse effects in patients receiving the standard therapy?”

Before the widespread use of experimental trials, clinicians attempted to answer such questions by generalizing from the experiences of individual patients to the population at large. Clinical judgement and reasoning were applied to reports of interesting cases. The concepts of variability among individuals and its sources were not formally addressed.

As the field of statistics, the “theoretical science or formal study of the inferential process, especially the planning and analysis of experiments, surveys, and observational studies.” (Piantadosi 2005). has developed in the twentieth century, clinical research has utilized statistical methods to provide formal accounting for sources of variability in patients’responses to treatment. The use of statistics allows clinical researchers to draw reasonable and accurate inferences from collected information and to make sound decisions in the presence of uncertainty. Mastery of statistical concepts can prevent numerous errors and biases in medical research.

Statistical reasoning is characterized by the following:

(Piantadosi, 2005)

• Establishing an objective framework for conducting an investigation
• Placing data and theory on an equal scientific footing
• Designing data production through experimentation
• Quantifying the influence of chance
• Estimating systematic and random effects
• Combining theory and data using formal methods

Carter, Scheaffer, and Marks (1986) stated that:

Clinical and statistical reasoning are both crucial to progress in medicine. Clinical researchers must generalize from the few to many and combine empirical evidence with theory. In both medical and statistical sciences, empirical knowledge is generated from observations and data. Medical theory is based upon established biology and hypotheses. Statistical theory is derived from mathematical and probabilistic models. (Piantadosi 2005), To establish a hypothesis requires both a theoretical basis in biology and statistical support for the hypothesis, based on the observed data and the theoretical statistical model.

What constitutes a ‘clinical trial’? Section  

An experiment is a series of observations made under conditions controlled by the scientist.

A clinical trial actually is an experiment testing medical treatments on human subjects. The clinical investigator controls factors that contribute to variability and bias such as the selection of subjects, application of the treatment, evaluation of outcome, and methods of analysis. The distinction of a clinical trial from other types of medical studies is the experimental nature of the trial and its occurrence in humans.

Design is the process or structure that isolates the factors of interest. Although the researcher designs a trial to control variability due to factors other than the treatment of interest, there is inherently larger variability in research involving humans than in a controlled laboratory situation.

The term “clinical trial” is preferred over “clinical experiment” because the latter may connote disrespect for the value of human life.

In what contexts are clinical trials used? Section  

Clinical trials are used to develop and test interventions in nearly all areas of medicine and public health. In many countries, approval for marketing new drugs hinges on efficacy and safety results from clinical trials. Similar requirements exist for the marketing of vaccines. The U.S. Food and Drug Administration (FDA) now requires manufacturers of new or high-risk medical devices to provide data demonstrating clinical safety and effectiveness. (Scott 2004). Surgical interventions pose unique challenges since surgical approaches are typically undertaken for patients with a good prognosis and may not be amenable to randomization or masking investigators and patients to the intervention, all conditions which can lead to biases. Clinical trials are useful for demonstrating efficacy and safety of various medical therapies, preventative measures and diagnostic procedures if the treatment can be applied uniformly and potential biases controlled.

In addition to testing novel therapies, clinical trials frequently are used to confirm findings from earlier studies. When the results of a study are surprising or contradict biological theory, a confirmatory trial may follow. Medical practice generally does not change based upon the results of one study. Design flaws, methodological errors, problems with study conduct, or analysis and reporting mistakes can render a clinical trial suspect. Hence, confirmation of results in a replicative study, or a trial extending the use of the therapy to a different population, is often warranted.

Clinical trials are time-consuming, labor-intensive, and expensive and require the cooperative effort of physicians, patients, nurses, data managers, methodologists, and statisticians. Patient recruitment can be difficult. Some multi-center (across institutions) clinical trials cost up to hundreds of million of dollars and take five years or more to complete. Prevention trials, conducted in healthy subjects to determine if treatments prevent the onset of disease, are important but the most cumbersome, lengthy, and expensive to conduct.

Many studies have a “window of opportunity” during which they are most feasible and will have the greatest impact on clinical practice. For comparative trials, the window usually exists relatively early in the development of a new therapy. If the treatment becomes widely accepted or discounted based on anecdotal experience, it may become impossible to formally test the efficacy of the procedure. Even when clinicians remain unconvinced of efficacy or relative safety, patient recruitment can become problematic.

Some important medical advances have been made without the formal methods of controlled clinical trials, i.e., without randomization, statistical design, and analysis. Examples include the use of vitamins, insulin, some antibiotics, and some vaccines.

Piantadosi (2005) gives the following requirements for a study based on a non-experimental comparative design to provide valid and convincing evidence:

• The treatment of interest must occur naturally
• The study subjects have to provide valid observations for the biological question
• The natural history of the disease with standard therapy, or in the absence of therapy, must be known
• The effect of the treatment must be large enough to overshadow random error and bias
• Evidence of efficacy must be consistent with biological knowledge

Examples of non-experimental designs that can yield convincing evidence of treatment efficacy can be found among epidemiological studies, historically-controlled trials, and from data mining.

Department of Statistics - Donald Bren School of Information & Computer Sciences

What is statistics.

Statistics is the science concerned with developing and studying methods for collecting, analyzing, interpreting and presenting empirical data. Statistics is a highly interdisciplinary field; research in statistics finds applicability in virtually all scientific fields and research questions in the various scientific fields motivate the development of new statistical methods and theory. In developing methods and studying the theory that underlies the methods statisticians draw on a variety of mathematical and computational tools.

Two fundamental ideas in the field of statistics are uncertainty and variation. There are many situations that we encounter in science (or more generally in life) in which the outcome is uncertain. In some cases the uncertainty is because the outcome in question is not determined yet (e.g., we may not know whether it will rain tomorrow) while in other cases the uncertainty is because although the outcome has been determined already we are not aware of it (e.g., we may not know whether we passed a particular exam).

Probability is a mathematical language used to discuss uncertain events and probability plays a key role in statistics. Any measurement or data collection effort is subject to a number of sources of variation. By this we mean that if the same measurement were repeated, then the answer would likely change. Statisticians attempt to understand and control (where possible) the sources of variation in any situation.

We encourage you to continue exploring our website to learn more about statistics, our academic programs, our students and faculty, as well as the cutting-edge research we are doing in the field.

• Core Members
• Affiliate Members
• Interdisciplinary Doctoral Program in Statistics
• Minor in Statistics and Data Science
• MicroMasters program in Statistics and Data Science
• Data Science and Machine Learning: Making Data-Driven Decisions
• Norbert Wiener Fellowship
• Stochastics and Statistics Seminar
• IDSS Distinguished Seminars
• IDSS Special Seminar
• SDSC Special Events
• Online events
• IDS.190 Topics in Bayesian Modeling and Computation
• Past Events
• LIDS & Stats Tea

Podcast: Data Nation

IDSS faculty and industry experts unpack how data can be used to lead, mislead, manipulate, and inform the public’s viewpoints and decisions. Season 2 has begun!

MicroMasters in Statistics and Data Science

Learn data science methods and tools, get hands-on training in data analysis and machine learning, and find opportunities in a growing field. Watch our latest informational webinar .

This 10-week online program covers statistics and Python foundations, machine learning, prediction, recommendation systems, and more.

Nonparametric Bayesian Statistics

Bayesian nonparametrics provides modeling solutions by replacing the finite-dimensional prior distributions of classical Bayesian analysis with infinite-dimensional stochastic processes.

Causal inference and applications to learning gene regulatory networks

Causal inference: Geometry of conditional independence structures for 3-node directed Gaussian graphical models.

Combinatorial learning with set functions

Learning problems that involve combinatorial objects are ubiquitous - they include the prediction of graphs, assignments, rankings, trees, groups of discrete labels or preferred sets of a user; the expression of prior structural knowledge for regularization, the identification of sets of important variables, or inference in discrete probabilistic models.

Online Learning

In this line of research, we develop strategies to optimize utility in dynamic environments in an optimal and efficient fashion.

Statistical and Computational Tradeoffs

Computational limitations of statistical problems have largely been ignored or simply overcome by ad hoc relaxations techniques.

MIT Statistics + Data Science Center Massachusetts Institute of Technology 77 Massachusetts Avenue Cambridge, MA 02139-4307 617-253-1764

• Accessibility
• Interdisciplinary PhD in Aero/Astro and Statistics
• Interdisciplinary PhD in Brain and Cognitive Sciences and Statistics
• Interdisciplinary PhD in Economics and Statistics
• Interdisciplinary PhD in Mathematics and Statistics
• Interdisciplinary PhD in Mechanical Engineering and Statistics
• Interdisciplinary PhD in Physics and Statistics
• Interdisciplinary PhD in Political Science and Statistics
• Interdisciplinary PhD in Social & Engineering Systems and Statistics
• LIDS & Stats Tea
• Spring 2023
• Spring 2022
• Spring 2021
• Fall – Spring 2020
• Fall 2019 – IDS.190 – Topics in Bayesian Modeling and Computation
• Fall 2019 – Spring 2019
• Fall 2018 and earlier

Statistics Made Easy

The Importance of Statistics in Psychology (With Examples)

The field of statistics is concerned with collecting, analyzing, interpreting, and presenting data.

In the field of psychology, statistics is important for the following reasons:

Reason 1 : Descriptive statistics allow psychologists to summarize data related to human performance, happiness, and other metrics.

Reason 2 : Regression models allow psychologists to quantify the relationship between variables related to human performance, happiness, and other metrics.

Reason 3 : Hypothesis tests allow psychologists to compare the effectiveness of different methods, techniques, and procedures on human performance, happiness, and other metrics.

In the rest of this article, we elaborate on each of these reasons.

Reason 1: Using Descriptive Statistics to Summarize Data

Descriptive statistics are used to describe data.

Psychologists often use descriptive statistics to summarize data related to individuals.

For example, an industrial-organizational psychologist might calculate the following descriptive statistics for individuals who work at a certain company:

• Overall satisfaction with salary (e.g. scale of 1-7)
• Overall satisfaction with workplace culture
• Overall satisfaction with working hours

Using these metrics, an I/O psychologist can gain a better understanding of how satisfied employees are at the company.

They can then use these metrics to inform the organization on areas that could use improvement to make the workplace a more enjoyable environment for the employees.

Reason 2: Using Regression Models to Quantify the Relationship Between Variables

Another way that statistics is used in psychology is in the form of regression models .

These are models that allow psychologists to quantify the relationship between one or more predictor variables and a response variable .

For example, a psychologist may have access to data on total hours spent exercising per day, total hours spent working per day, and overall  happiness (e.g. scale of 0-100) of individuals.

They might then build the following multiple linear regression model:

Happiness = 76.4 + 9.3(hours spent exercising per day) – 0.4(hours spent working per day)

Here’s how to interpret the regression coefficients in this model:

• For each additional hour spent exercising per day, overall happiness increases by an average of 9.3 points (assuming hours spent working is held constant).
• For each additional hour spent working per day, overall happiness decreases by an average of 0.4 points (assuming hours spent exercising is held constant).

Using this model, a psychologist can quickly understand that more time spent exercising is associated with increased overall happiness and more time spent working is associated with lower overall happiness.

They can also quantify exactly how much exercise and working affect overall happiness.

Reason 3: Using Hypothesis Tests to Compare Methods

Another way that statistics is used in psychology is in the form of hypothesis tests .

These are tests that psychologists can use to determine if there is a statistical significance between different methods, techniques, or procedures.

For example, suppose a sports psychologist believes that a new workout method is able to increase the mental well-being of college basketball players. To test this, he may measure the well-being (e.g. scale of 1-7) of 40 players before and after implementing the new workout method for one month.

He can then perform a paired samples t- test using the following hypotheses:

• H 0 : μ after = μ before (the mean well-being is the same before and after using the method)
• H A : μ after > μ before (the mean well-being is greater after using the method)

If the p-value of the test is less than some significance level (e.g. α = .05), then he can reject the null hypothesis and conclude that the new method leads to increased well-being among players

Note : This is just one example of a hypothesis test that is used in psychology. Other common tests include a one sample t-test , two sample t-test , one-way ANOVA , and two-way ANOVA .

Additional Resources

The following articles explain the importance of statistics in other fields:

The Importance of Statistics in Research The Importance of Statistics in Healthcare The Importance of Statistics in Business The Importance of Statistics in Economics The Importance of Statistics in Education

Featured Posts

Hey there. My name is Zach Bobbitt. I have a Masters of Science degree in Applied Statistics and I’ve worked on machine learning algorithms for professional businesses in both healthcare and retail. I’m passionate about statistics, machine learning, and data visualization and I created Statology to be a resource for both students and teachers alike.  My goal with this site is to help you learn statistics through using simple terms, plenty of real-world examples, and helpful illustrations.

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Join the Statology Community

Sign up to receive Statology's exclusive study resource: 100 practice problems with step-by-step solutions. Plus, get our latest insights, tutorials, and data analysis tips straight to your inbox!

By subscribing you accept Statology's Privacy Policy.

• Search Search Please fill out this field.

What Is Statistics?

Understanding statistics.

• Descriptive & Inferential Statistics

Mean, Median, and Mode

Understanding statistical data.

• Levels of Measurement
• Sampling Techniques

Uses of Statistics

The bottom line.

• Corporate Finance
• Financial Analysis

Statistics: Definition, Types, and Importance

Katrina Ávila Munichiello is an experienced editor, writer, fact-checker, and proofreader with more than fourteen years of experience working with print and online publications.

Statistics is a branch of applied mathematics that involves the collection, description, analysis, and inference of conclusions from quantitative data. The mathematical theories behind statistics rely heavily on differential and integral calculus, linear algebra, and probability theory.

People who do statistics are referred to as statisticians. They’re particularly concerned with determining how to draw reliable conclusions about large groups and general events from the behavior and other observable characteristics of small samples. These small samples represent a portion of the large group or a limited number of instances of a general phenomenon.

Key Takeaways

• Statistics is the study and manipulation of data, including ways to gather, review, analyze, and draw conclusions from data.
• The two major areas of statistics are descriptive and inferential statistics.
• Statistics can be communicated at different levels ranging from non-numerical descriptor (nominal-level) to numerical in reference to a zero-point (ratio-level).
• Several sampling techniques can be used to compile statistical data, including simple random, systematic, stratified, or cluster sampling.
• Statistics are present in almost every department of every company and are an integral part of investing.

Dennis Madamba / Investopedia

Statistics are used in virtually all scientific disciplines, such as the physical and social sciences as well as in business, the humanities, government, and manufacturing. Statistics is fundamentally a branch of applied mathematics that developed from the application of mathematical tools, including calculus and linear algebra, to probability theory.

In practice, statistics is the idea that we can learn about the properties of large sets of objects or events (a population ) by studying the characteristics of a smaller number of similar objects or events (a sample ). Gathering comprehensive data about an entire population is too costly, difficult, or impossible in many cases, so statistics start with a sample that can be conveniently or affordably observed.

Statisticians measure and gather data about the individuals or elements of a sample, then they analyze this data to generate descriptive statistics. They can then use these observed characteristics of the sample data, which are properly called “statistics,” to make inferences or educated guesses about the unmeasured characteristics of the broader population, known as the parameters.

Statistics informally dates back centuries. An early record of correspondence between French mathematicians Pierre de Fermat and Blaise Pascal in 1654 is often cited as an early example of statistical probability analysis.

Descriptive and Inferential Statistics

The two major areas of statistics are known as descriptive statistics , which describes the properties of sample and population data, and inferential statistics, which uses those properties to test hypotheses and draw conclusions. Descriptive statistics include mean (average), variance, skewness , and kurtosis . Inferential statistics include linear regression analysis, analysis of variance (ANOVA), logit/Probit models, and null hypothesis testing.

Descriptive Statistics

Descriptive statistics mostly focus on the central tendency, variability, and distribution of sample data. Central tendency means the estimate of the characteristics, a typical element of a sample or population. It includes descriptive statistics such as mean , median , and mode .

Variability refers to a set of statistics that show how much difference there is among the elements of a sample or population along the characteristics measured. It includes metrics such as range, variance , and standard deviation .

The distribution refers to the overall “shape” of the data, which can be depicted on a chart such as a histogram or a dot plot, and includes properties such as the probability distribution function, skewness, and kurtosis. Descriptive statistics can also describe differences between observed characteristics of the elements of a data set. They can help us understand the collective properties of the elements of a data sample and form the basis for testing hypotheses and making predictions using inferential statistics.

Inferential Statistics

Inferential statistics is a tool that statisticians use to draw conclusions about the characteristics of a population, drawn from the characteristics of a sample, and to determine how certain they can be of the reliability of those conclusions. Based on the sample size and distribution, statisticians can calculate the probability that statistics, which measure the central tendency, variability, distribution, and relationships between characteristics within a data sample, provide an accurate picture of the corresponding parameters of the whole population from which the sample is drawn.

Inferential statistics are used to make generalizations about large groups, such as estimating average demand for a product by surveying a sample of consumers’ buying habits or attempting to predict future events. This might mean projecting the future return of a security or asset class based on returns in a sample period.

Regression  analysis is a widely used technique of statistical inference used to determine the strength and nature of the relationship (the correlation) between a dependent variable and one or more explanatory (independent) variables. The output of a regression model is often analyzed for statistical significance, which refers to the claim that a result from findings generated by testing or experimentation is not likely to have occurred randomly or by chance. It’s likely to be attributable to a specific cause elucidated by the data.

Having statistical significance is important for academic disciplines or practitioners that rely heavily on analyzing data and research.

The terms “mean,” “median,” and “mode” fall under the umbrella of central tendency. They describe an element that’s typical in a given sample group. You can find the mean descriptor by adding the numbers in the group and dividing the result by the number of data set observations.

The middle number in the set is the median. Half of all included numbers are higher than the median, and half are lesser. The median home value in a neighborhood would be $350,000 if five homes were located there and valued at $500,000, $400,000, $350,000, $325,000, and $300,000. Two values are higher, and two are lower.

Mode identifies the number that falls between the highest and lowest values. It appears most frequently in the data set.

The root of statistics is driven by variables. A variable is a data set that can be counted that marks a characteristic or attribute of an item. For example, a car can have variables such as make, model, year, mileage, color, or condition. By combining the variables across a set of data, such as the colors of all cars in a given parking lot, statistics allows us to better understand trends and outcomes.

There are two main types of variables:

First, qualitative variables are specific attributes that are often non-numeric. Many of the examples given in the car example are qualitative. Other examples of qualitative variables in statistics are gender, eye color, or city of birth. Qualitative data is most often used to determine what percentage of an outcome occurs for any given qualitative variable. Qualitative analysis often does not rely on numbers. For example, trying to determine what percentage of women own a business analyzes qualitative data.

The second type of variable in statistics is quantitative variables. Quantitative variables are studied numerically and only have weight when they’re about a non-numerical descriptor. Similar to quantitative analysis, this information is rooted in numbers. In the car example above, the mileage driven is a quantitative variable, but the number 60,000 holds no value unless it is understood that is the total number of miles driven.

Quantitative variables can be further broken into two categories. First, discrete variables have limitations in statistics and infer that there are gaps between potential discrete variable values. The number of points scored in a football game is a discrete variable because:

• There can be no decimals.
• It is impossible for a team to score only one point.

Statistics also makes use of continuous quantitative variables. These values run along a scale. Discrete values have limitations, but continuous variables are often measured into decimals. Any value within possible limits can be obtained when measuring the height of the football players, and the heights can be measured down to 1/16th of an inch, if not further.

Statisticians can hold various titles and positions within a company. The average total compensation for a statistician with one to three years of experience was $81,885 as of December 2023. This increased to $109,288 with 15 years of experience.

Statistical Levels of Measurement

There are several resulting levels of measurement after analyzing variables and outcomes. Statistics can quantify outcomes in four ways.

Nominal-level Measurement

There’s no numerical or quantitative value, and qualities are not ranked. Nominal-level measurements are instead simply labels or categories assigned to other variables. It’s easiest to think of nominal-level measurements as non-numerical facts about a variable.

Example : The name of the president elected in 2020 was Joseph Robinette Biden Jr.

Ordinal-level Measurement

Outcomes can be arranged in an order, but all data values have the same value or weight. Although they’re numerical, ordinal-level measurements can’t be subtracted against each other in statistics because only the position of the data point matters. Ordinal levels are often incorporated into nonparametric statistics and compared against the total variable group.

Example : American Fred Kerley was the second-fastest man at the 2020 Tokyo Olympics based on 100-meter sprint times.

Interval-level Measurement

Outcomes can be arranged in order, but differences between data values may now have meaning. Two data points are often used to compare the passing of time or changing conditions within a data set. There is often no “starting point” for the range of data values, and calendar dates or temperatures may not have a meaningful intrinsic zero value.

Example : Inflation hit 8.6% in May 2022. The last time inflation was this high was in December 1981 .

Ratio-level Measurement

Outcomes can be arranged in order, and differences between data values now have meaning. But there’s a starting point or “zero value” that can be used to further provide value to a statistical value. The ratio between data values has meaning, including its distance away from zero.

Example : The lowest meteorological temperature recorded was -128.6 degrees Fahrenheit in Antarctica.

Statistics Sampling Techniques

It would often not be possible to gather data from every data point within a population to gather statistical information. Statistics relies instead on different sampling techniques to create a representative subset of the population that’s easier to analyze. In statistics, there are several primary types of sampling.

Simple Random Sampling

Simple random sampling calls for every member within the population to have an equal chance of being selected for analysis. The entire population is used as the basis for sampling, and any random generator based on chance can select the sample items. For example, 100 individuals are lined up and 10 are chosen at random.

Systemic Sampling

Systematic sampling calls for a random sample as well, but its technique is slightly modified to make it easier to conduct. A single random number is generated, and individuals are then selected at a specified regular interval until the sample size is complete. For example, 100 individuals are lined up and numbered. The seventh individual is selected for the sample, followed by every subsequent ninth individual, until 10 sample items have been selected.

Stratified Sampling

Stratified sampling calls for more control over your sample. The population is divided into subgroups based on similar characteristics. Then you calculate how many people from each subgroup would represent the entire population. For example, 100 individuals are grouped by gender and race. Then a sample from each subgroup is taken in proportion to how representative that subgroup is of the population.

Cluster Sampling

Cluster sampling calls for subgroups as well, but each subgroup should be representative of the population. The entire subgroup is randomly selected instead of randomly selecting individuals within a subgroup.

Not sure which Major League Baseball player should have won Most Valuable Player last year? Statistics, often used to determine value, is often cited when the award for best player is awarded. Statistics can include batting average, number of home runs hit, and stolen bases.

Statistics is prominent in finance, investing, business, and in the world. Much of the information you see and the data you’re given is derived from statistics, which are used in all facets of a business.

• Statistics in investing include average trading volume, 52-week low, 52-week high, beta, and correlation between asset classes or securities.
• Statistics in economics include gross domestic product (GDP), unemployment, consumer pricing, inflation, and other economic growth metrics.
• Statistics in marketing include conversion rates, click-through rates, search quantities, and social media metrics.
• Statistics in accounting include liquidity, solvency, and profitability metrics across time.
• Statistics in information technology include bandwidth, network capabilities, and hardware logistics.
• Statistics in human resources include employee turnover, employee satisfaction, and average compensation relative to the market.

Why Is Statistics Important?

Statistics provide the information to educate how things work. They’re used to conduct research, evaluate outcomes, develop critical thinking, and make informed decisions. Statistics can be used to inquire about almost any field of study to investigate why things happen, when they occur, and whether reoccurrence is predictable.

What’s the Difference Between Descriptive and Inferential Statistics?

Descriptive statistics are used to describe or summarize the characteristics of a sample or data set, such as a variable’s mean, standard deviation, or frequency. Inferential statistics employ any number of techniques to relate variables in a data set to one another. An example would be using correlation or regression analysis. These can then be used to estimate forecasts or infer causality.

Who Uses Statistics?

Statistics are used widely across an array of applications and professions. Statistics are done whenever data are collected and analyzed. This can range from government agencies to academic research to analyzing investments.

How Are Statistics Used in Economics and Finance?

Economists collect and look at all sorts of data ranging from consumer spending to housing starts to inflation to GDP growth. In finance, analysts and investors collect data about companies, industries, sentiment, and market data on price and volume. The use of inferential statistics in these fields is known as econometrics. Several important financial models, from the capital asset pricing model (CAPM) to modern portfolio theory (MPT) and the Black-Scholes options pricing model, rely on statistical inference.

Statistics is the practice of analyzing pieces of information that might seem conflicting or unrelated at first glance and on the surface. It can lead to a solid career as a statistician, but it can also be a handy metric in everyday life—perhaps when you’re analyzing the odds that your favorite team will win the Super Bowl before you place a bet, gauging the viability of an investment, or determining whether you’re being comparatively overcharged for a product or service.

Encyclopœdia Britannica. “ Probability and Statistics .”

Coursera. “ How Much Do Statisticians Make? Your 2024 Salary Guide .”

Olympics. “ Tokyo 2020: Athletics Men’s 100m Results .”

U.S. Bureau of Labor Statistics. “ Consumer Price Index .”

Arizona State University, World Meteorological Organization’s World Weather & Climate Extremes Archive. “ World: Lowest Temperature .”

Baseball Reference. “ MLB Most Valuable Player MVP Award Winners .”

• Terms of Service
• Editorial Policy
• Privacy Policy
• Your Privacy Choices

Explaining research performance: investigating the importance of motivation

• Original Paper
• Open access
• Published: 23 May 2024
• Volume 4 , article number  105 , ( 2024 )

Cite this article

You have full access to this open access article

• Silje Marie Svartefoss   ORCID: orcid.org/0000-0001-5072-1293 1   nAff4 ,
• Jens Jungblut 2 ,
• Dag W. Aksnes 1 ,
• Kristoffer Kolltveit 2 &
• Thed van Leeuwen 3  

421 Accesses

6 Altmetric

Explore all metrics

In this article, we study the motivation and performance of researchers. More specifically, we investigate what motivates researchers across different research fields and countries and how this motivation influences their research performance. The basis for our study is a large-N survey of economists, cardiologists, and physicists in Denmark, Norway, Sweden, the Netherlands, and the UK. The analysis shows that researchers are primarily motivated by scientific curiosity and practical application and less so by career considerations. There are limited differences across fields and countries, suggesting that the mix of motivational aspects has a common academic core less influenced by disciplinary standards or different national environments. Linking motivational factors to research performance, through bibliometric data on publication productivity and citation impact, our data show that those driven by practical application aspects of motivation have a higher probability for high productivity. Being driven by career considerations also increases productivity but only to a certain extent before it starts having a detrimental effect.

Similar content being viewed by others

Are the strategic research agendas of researchers in the social sciences determinants of research productivity?

Research incentives and research output.

Exploring the determinants of research performance for early-career researchers: a literature review

Avoid common mistakes on your manuscript.

Introduction

Motivation and abilities are known to be as important factors in explaining employees’ job performance of employees (Van Iddekinge et al. 2018 ), and in the vast scientific literature on motivation, it is common to differentiate between intrinsic and extrinsic motivation factors (Ryan and Deci 2000 ). In this context, path-breaking individuals are said to often be intrinsically motivated (Jindal-Snape and Snape 2006 ; Thomas and Nedeva 2012 ; Vallerand et al. 1992 ), and it has been found that the importance of these of types of motivations differs across occupations and career stages (Duarte and Lopes 2018 ).

In this article, we address the issue of motivation for one specific occupation, namely: researchers working at universities. Specifically, we investigate what motivates researchers across fields and countries (RQ1) and how this motivation is linked to their research performance (RQ2). The question of why people are motivated to do their jobs is interesting to address in an academic context, where work is usually harder to control, and individuals tend to have a lot of much freedom in structuring their work. Moreover, there have been indications that academics possess an especially high level of motivation for their tasks that is not driven by a search for external rewards but by an intrinsic satisfaction from academic work (Evans and Meyer 2003 ; Leslie 2002 ). At the same time, elements of researchers’ performance are measurable through indicators of their publication activity: their productivity through the number of outputs they produce and the impact of their research through the number of citations their publications receive (Aksnes and Sivertsen 2019 ; Wilsdon et al. 2015 ).

Elevating research performance is high on the agenda of many research organisations (Hazelkorn 2015 ). How such performance may be linked to individuals’ motivational aspects has received little attention. Thus, a better understanding of this interrelation may be relevant for developing institutional strategies to foster environments that promote high-quality research and research productivity.

Previous qualitative research has shown that scientists are mainly intrinsically motivated (Jindal-Snape and Snape 2006 ). Other survey-based contributions suggest that there can be differences in motivations across disciplines (Atta-Owusu and Fitjar 2021 ; Lam 2011 ). Furthermore, the performance of individual scientists has been shown to be highly skewed in terms of publication productivity and citation rates (Larivière et al. 2010 ; Ruiz-Castillo and Costas 2014 ). There is a large body of literature explaining these differences. Some focus on national and institutional funding schemes (Hammarfelt and de Rijcke 2015 ; Melguizo and Strober 2007 ) and others on the research environment, such as the presence of research groups and international collaboration (Jeong et al. 2014 ), while many studies address the role of academic rank, age, and gender (see e.g. Baccini et al. 2014 ; Rørstad and Aksnes 2015 ). Until recently, less emphasis has been placed on the impact of researchers’ motivation. Some studies have found that different types of motivations drive high levels of research performance (see e.g. Horodnic and Zaiţ 2015 ; Ryan and Berbegal-Mirabent 2016 ). However, researchers are only starting to understand how this internal drive relates to research performance.

While some of the prior research on the impact of motivation depends on self-reported research performance evaluations (Ryan 2014 ), the present article combines survey responses with actual bibliometric data. To investigate variation in research motivation across scientific fields and countries, we draw on a large-N survey of economists, cardiologists, and physicists in Denmark, Norway, Sweden, the Netherlands, and the UK. To investigate how this motivation is linked to their research performance, we map the survey respondents’ publication and citation data from the Web of Science (WoS).

This article is organised as follows. First, we present relevant literature on research performance and motivation. Next, the scientific fields and countries are then presented before elaborating on our methodology. In the empirical analysis, we investigate variations in motivation across fields, gender, age, and academic position and then relate motivation to publications and citations as our two measures of research performance. In the concluding section, we discuss our findings and implications for national decision-makers and individual researchers.

Motivation and research performance

As noted above, the concepts of intrinsic and extrinsic motivation play an important role in the literature on motivation and performance. Here, intrinsic motivation refers to doing something for its inherent satisfaction rather than for some separable consequence. Extrinsic motivation refers to doing something because it leads to a separable outcome (Ryan and Deci 2000 ).

Some studies have found that scientists are mainly intrinsically motivated (Jindal-Snape and Snape 2006 ; Lounsbury et al. 2012 ). Research interests, curiosity, and a desire to contribute to new knowledge are examples of such motivational factors. Intrinsic motives have also been shown to be crucial when people select research as a career choice (Roach and Sauermann 2010 ). Nevertheless, scientists are also motivated by extrinsic factors. Several European countries have adopted performance-based research funding systems (Zacharewicz et al. 2019 ). In these systems, researchers do not receive direct financial bonuses when they publish, although such practices may occur at local levels (Stephan et al. 2017 ). Therefore, extrinsic motivation for such researchers may include salary increases, peer recognitions, promotion, or expanded access to research resources (Lam 2011 ). According to Tien and Blackburn ( 1996 ), both types of motivations operate simultaneously, and their importance vary and may depend on the individual’s circumstances, personal situation, and values.

The extent to which different kinds of motivations play a role in scientists’ performance has been investigated in several studies. In these studies, bibliometric indicators based on the number of publications are typically used as outcome measures. Such indicators play a critical role in various contexts in the research system (Wilsdon et al. 2015 ), although it has also been pointed out that individuals can have different motivations to publish (Hangel and Schmidt-Pfister 2017 ).

Based on a survey of Romanian economics and business administration academics combined with bibliometric data, Horodnic and Zait ( 2015 ) found that intrinsic motivation was positively correlated with research productivity, while extrinsic motivation was negatively correlated. Their interpretations of the results are that researchers motivated by scientific interest are more productive, while researchers motivated by extrinsic forces will shift their focus to more financially profitable activities. Similarly, based on the observation that professors continue to publish even after they have been promoted to full professor, Finkelstein ( 1984 ) concluded that intrinsic rather than extrinsic motivational factors have a decisive role regarding the productivity of academics.

Drawing on a survey of 405 research scientists working in biological, chemical, and biomedical research departments in UK universities, Ryan ( 2014 ) found that (self-reported) variations in research performance can be explained by instrumental motivation based on financial incentives and internal motivation based on the individual’s view of themselves (traits, competencies, and values). In the study, instrumental motivation was found to have a negative impact on research performance: As the desire for financial rewards increase, the level of research performance decreases. In other words, researchers mainly motivated by money will be less productive and effective in their research. Contrarily, internal motivation was found to have a positive impact on research performance. This was explained by highlighting that researchers motivated by their self-concept set internal standards that become a reference point that reinforces perceptions of competency in their environments.

Nevertheless, it has also been argued that intrinsic and extrinsic motivations for publishing are intertwined (Ma 2019 ). According to Tien and Blackburn ( 1996 ), research productivity is neither purely intrinsically nor purely extrinsically motivated. Publication activity is often a result of research, which may be intrinsically motivated or motivated by extrinsic factors such as a wish for promotion, where the number of publications is often a part of the assessment (Cruz-Castro and Sanz-Menendez 2021 ; Tien 2000 , 2008 ).

The negative relationship between external/instrumental motivation and performance and the positive relationship between internal/self-concept motivation and performance are underlined by Ryan and Berbegal-Mirabent ( 2016 ). Drawing on a fuzzy set qualitative comparative analysis of a random sampling of 300 of the original respondents from Ryan ( 2014 ), they find that scientists working towards the standards and values they identify with, combined with a lack of concern for instrumental rewards, contribute to higher levels of research performance.

Based on the above, this article will address two research questions concerning different forms of motivation and the relationship between motivation and research performance.

How does the motivation of researchers vary across fields and countries?

How do different types of motivations affect research performance?

In this study, the roles of three different motivational factors are analysed. These are scientific curiosity, practical and societal applications, and career progress. The study aims to assess the role of these specific motivational factors and not the intrinsic-extrinsic distinction more generally. Of the three factors, scientific curiosity most strongly relates to intrinsic motivation; practical and societal applications also entail strong intrinsic aspects. On the other hand, career progress is linked to extrinsic motivation.

In addition to variation in researchers’ motivations by field and country, we consider differences in relation to age, position and gender. Additionally, when investigating how motivation relates to scientific performance we control for the influence of age, gender, country and funding. These are dimensions where differences might be found in motivational factors given that scientific performance, particularly publication productivity, has been shown to differ along these dimensions (Rørstad and Aksnes 2015 ).

Research context: three fields, five countries

To address the research question about potential differences across fields and countries, the study is based on a sample consisting of researchers in three different fields (cardiology, economics, and physics) and five countries (Denmark, Norway, Sweden, the Netherlands, and the UK). Below, we describe this research context in greater detail.

The fields represent three different domains of science: medicine, social sciences, and the natural sciences, where different motivational factors may be at play. This means that the fields cover three main areas of scientific investigations: the understanding of the world, the functioning of the human body, and societies and their functions. The societal role and mission of the fields also differ. While a primary aim of cardiology research and practice is to reduce the burden of cardiovascular disease, physics research may drive technology advancements, which impacts society. Economics research may contribute to more effective use of limited resources and the management of people, businesses, markets, and governments. In addition, the fields also differ in publication patterns (Piro et al. 2013 ). The average number of publications per researcher is generally higher in cardiology and physics than in economics (Piro et al. 2013 ). Moreover, cardiologists and physicists mainly publish in international scientific journals (Moed 2005 ; Van Leeuwen 2013 ). In economics, researchers also tend to publish books, chapters, and articles in national languages, in addition to international journal articles (Aksnes and Sivertsen 2019 ; van Leeuwen et al. 2016 ).

We sampled the countries with a twofold aim. On the one hand, we wanted to have countries that are comparable so that differences in the development of the science systems, working conditions, or funding availability would not be too large. On the other hand, we also wanted to assure variation among the countries regarding these relevant framework conditions to ensure that our findings are not driven by a specific contextual condition.

The five countries in the study are all located in the northwestern part of Europe, with science systems that are foremost funded by block grant funding from the national governments (unlike, for example, the US, where research grants by national funding agencies are the most important funding mechanism) (Lepori et al. 2023 ).

In all five countries, the missions of the universities are composed of a blend of education, research, and outreach. Furthermore, the science systems in Norway, Denmark, Sweden, and the Netherlands have a relatively strong orientation towards the Anglo-Saxon world in the sense that publishing in the national language still exists, but publishing in English in internationally oriented journals in which English is the language of publications is the norm (Kulczycki et al. 2018 ). These framework conditions ensure that those working in the five countries have somewhat similar missions to fulfil in their professions while also belonging to a common mainly Anglophone science system.

However, in Norway, Denmark, Sweden, and the Netherlands, research findings in some social sciences, law, and the humanities are still oriented on publishing in various languages. Hence, we avoided selecting the humanities field for this study due to a potential issue with cross-country comparability (Sivertsen 2019 ; Sivertsen and Van Leeuwen 2014 ; Van Leeuwen 2013 ).

Finally, the chosen countries vary regarding their level of university autonomy. When combining the scores for organisational, financial, staffing, and academic autonomy presented in the latest University Autonomy in Europe Scorecard presented by the European University Association (EUA), the UK, the Netherlands, and Denmark have higher levels of autonomy compared to Norway and Sweden, with Swedish universities having less autonomy than their Norwegian counterparts (Pruvot et al. 2023 ). This variation is relevant for our study, as it ensures that our findings are not driven by response from a higher education system with especially high or low autonomy, which can influence the motivation and satisfaction of academics working in it (Daumiller et al. 2020 ).

Data and methods

The data used in this article are a combination of survey data and bibliometric data retrieved from the WoS. The WoS database was chosen for this study due to its comprehensive coverage of research literature across all disciplines, encompassing the three specific research areas under analysis. Additionally, the WoS database is well-suited for bibliometric analyses, offering citation counts essential for this study.

Two approaches were used to identify the sample for the survey. Initially, a bibliometric analysis of the WoS using journal categories (‘Cardiac & cardiovascular systems’, ‘Economics’, and ‘Physics’) enabled the identification of key institutions with a minimum number of publications within these journal categories. Following this, relevant organisational units and researchers within these units were identified through available information on the units’ webpages. Included were employees in relevant academic positions (tenured academic personnel, post-docs, and researchers, but not PhD students, adjunct positions, guest researchers, or administrative and technical personnel).

Second, based on the WoS data, people were added to this initial sample if they had a minimum number of publications within the field and belonged to any of the selected institutions, regardless of unit affiliation. For economics, the minimum was five publications within the selected period (2011–2016). For cardiology and physics, where the individual publication productivity is higher, the minimum was 10 publications within the same period. The selection of the minimum publication criteria was based on an analysis of publication outputs in these fields between 2011 and 2016. The thresholds were applied to include individuals who are more actively engaged in research while excluding those with more peripheral involvement. The higher thresholds for cardiology and physics reflect the greater frequency of publications (and co-authorship) observed in these fields.

The benefit of this dual-approach strategy to sampling is that we obtain a more comprehensive sample: the full scope of researchers within a unit and the full scope of researchers that publish within the relevant fields. Overall, 59% of the sample were identified through staff lists and 41% through the second step involving WoS data.

The survey data were collected through an online questionnaire first sent out in October 2017 and closed in December 2018. In this period, several reminders were sent to increase the response rate. Overall, the survey had a response rate of 26.1% ( N  = 2,587 replies). There were only minor variations in response rates between scientific fields; the variations were larger between countries. Tables  1 and 2 provide an overview of the response rate by country and field.

Operationalisation of motivation

Motivation was measured by a question in the survey asking respondents what motivates or inspires them to conduct research, of which three dimensions are analysed in the present paper. The two first answer categories were related to intrinsic motivation (‘Curiosity/scientific discovery/understanding the world’ and ‘Application/practical aims/creating a better society’). The third answer category was more related to extrinsic motivation (‘Progress in my career [e.g. tenure/permanent position, higher salary, more interesting/independent work]’). Appendix Table A1 displays the distribution of respondents and the mean value and standard deviation for each item.

These three different aspects of motivation do not measure the same phenomenon but seem to capture different aspects of motivation (see Pearson’s correlation coefficients in Appendix Table A2 ). There is no correlation between curiosity/scientific discovery, career progress, and practical application. However, there is a weak but significant positive correlation between career progress and practical application. These findings indicate that those motivated by career considerations to some degrees also are motivated by practical application.

In addition to investigating how researchers’ motivation varies by field and country, we consider the differences in relation to age, position and gender as well. Field of science differentiates between economics, cardiology, physics, and other fields. The country variables differentiate between the five countries. Age is a nine-category variable. The position variable differentiates between full professors, associate professors, and assistant professors. The gender variable has two categories (male or female). For descriptive statistics on these additional variables, see Appendix Table A3 .

Publication productivity and citation impact

To analyse the respondents’ bibliometric performance, the Centre for Science and Technology Studies (CWTS) in-house WoS database was used. We identified the publication output of each respondent during 2011–2017 (limited to regular articles, reviews, and letters). For 16% of the respondents, no publications were identified in the database. These individuals had apparently not published in international journals covered by the database. However, in some cases, the lack of publications may be due to identification problems (e.g. change of names). Therefore, we decided not to include the latter respondents in the analysis.

Two main performance measures were calculated: publication productivity and citation impact. As an indicator of productivity, we counted the number of publications for each individual (as author or co-author) during the period. To analyse the citation impact, a composite measure using three different indicators was used: total number of citations (total citations counts for all articles they have contributed to during the period, counting citations up to and including 2017), normalised citation score (MNCS), and proportion of publications among the 10% most cited articles in their fields (Waltman and Schreiber 2013 ). Here, the MNCS is an indicator for which the citation count of each article is normalised by subject, article type, and year, where 1.00 corresponds to the world average (Waltman et al. 2011 ). Based on these data, averages for the total publication output of each respondent were calculated. By using three different indicators, we can avoid biases or limitations attached to each of them. For example, using the MNCS, a respondent with only one publication would appear as a high impact researcher if this article was highly cited. However, when considering the additional indicator, total citation counts, this individual would usually perform less well.

The bibliometric scores were skewedly distributed among the respondents. Rather than using the absolute numbers, in this paper, we have classified the respondents into three groups according to their scores on the indicators. Here, we have used percentile rank classes (tertiles). Percentile statistics are increasingly applied in bibliometrics (Bornmann et al. 2013 ; Waltman and Schreiber 2013 ) due to the presence of outliers and long tails, which characterise both productivity and citation distributions.

As the fields analysed have different publication patterns, the respondents within each field were ranked according to their scores on the indicators, and their percentile rank was determined. For the productivity measure, this means that there are three groups that are equal in terms of number of individuals included: 1: Low productivity (the group with the lowest publication numbers, 0–33 percentile), 2: Medium productivity (33–67 percentile), and 3: High productivity (67–100 percentile). For the citation impact measure, we conducted a similar percentile analysis for each of the three composite indicators. Then everyone was assigned to one of the three percentile groups based on their average score: 1: Low citation impact (the group with lowest citation impact, 0–33 percentile), 2: Medium citation impact (33–67 percentile), and 3: High citation impact (67–100 percentile), cf. Table  3 . Although it might be argued that the application of tertile groups rather than absolute numbers leads to a loss of information, the advantage is that the results are not influenced by extreme values and may be easier to interpret.

Via this approach, we can analyse the two important dimensions of the respondents’ performance. However, it should be noted that the WoS database does not cover the publication output of the fields equally. Generally, physics and cardiology are very well covered, while the coverage of economics is somewhat lower due to different publication practices (Aksnes and Sivertsen 2019 ). This problem is accounted for in our study by ranking the respondents in each field separately, as described above. In addition, not all respondents may have been active researchers during the entire 2011–2017 period, which we have not adjusted for. Despite these limitations, the analysis provides interesting information on the bibliometric performance of the respondents at an aggregated level.

Regression analysis

To analyse the relationship between motivation and performance, we apply multinomial logistic regression rather then ordered logistic regression because we assume that the odds for respondents belonging in each category of the dependent variables are not equal (Hilbe 2017 ). The implication of this choice of model is that the model tests the probability of respondents being in one category compared to another (Hilbe 2017 ). This means that a reference or baseline category must be selected for each of the dependent variables (productivity and citation impact). Furthermore, the coefficient estimates show how the probability of being in one of the other categories decreases or increases compared to being in the reference category.

For this analysis, we selected the medium performers as the reference or baseline category for both our dependent variables. This enables us to evaluate how the independent variables affect the probability of being in the low performers group compared to the medium performers and the high performers compared to the medium performers.

To evaluate model fit, we started with a baseline model where only types of motivations were included as independent variables. Subsequently, the additional variables were introduced into the model, and based on measures for model fit (Pseudo R 2 , -2LL, and Akaike Information Criterion (AIC)), we concluded that the model with all additional variables included provides the best fit to the data for both the dependent variables (see Appendix Tables A5 and A6 ). Additional control variables include age, gender, country, and funding. We include these variables as controls to obtain robust effects of motivation and not effects driven by other underlying factors. The type of funding was measured by variables where the respondent answered the following question: ‘How has your research been funded the last five years?’ The funding variable initially consisted of four categories: ‘No source’, ‘Minor source’, ‘Moderate source’, and ‘Major source’. In this analysis, we have combined ‘No source’ and ‘Minor source’ into one category (0) and ‘Moderate source’ and ‘Major source’ into another category (1). Descriptive statistics for the funding variables are available in Appendix Table A4 . We do not control for the influence of field due to how the scientific performance variables are operationalised, the field normalisation implies that there are no variations across fields. We also do not control for position, as this variable is highly correlated with age, and we are therefore unable to include these two variables in the same model.

The motivation of researchers

In the empirical analysis, we first investigate variation in motivation and then relate it to publications and citations as our two measures of research performance.

As Fig.  1 shows, the respondents are mainly driven by curiosity and the wish to make scientific discoveries. This is by far the most important motivation. Practical application is also an important source of motivation, while making career progress is not identified as being very important.

Motivation of researchers– percentage

As Table  4 shows, at the level of fields, there are no large differences, and the motivational profiles are relatively similar. However, physicists tend to view practical application as somewhat less important than cardiologists and economists. Moreover, career progress is emphasised most by economists. Furthermore, as table 5 shows, there are some differences in motivation between countries. For curiosity/scientific discovery and practical application, the variations across countries are minor, but researchers in Denmark tend to view career progress as somewhat more important than researchers in the other countries.

Furthermore, as table 6 shows, women seem to view practical application and career progress as a more important motivation than men; these differences are also significant. Similar gender disparities have also been reported in a previous study (Zhang et al. 2021 ).

There are also some differences in motivation across the additional variables worth mentioning, as Table  7 shows. Unsurprisingly, perhaps, there is a significant moderate negative correlation between age, position, and career progress. This means that the importance of career progress as a motivation seems to decrease with increased age or a move up the position hierarchy.

In the second part of the analysis, we relate motivation to research performance. We first investigate publications and productivity using the percentile groups. Here, we present the results we use using predicted probabilities because they are more easily interpretable than coefficient estimates. For the model with productivity percentile groups as the dependent variable, the estimates for career progress were negative when comparing the medium productivity group to the high productivity group and the medium productivity group to the low productivity group. This result indicates that the probability of being in the high and low productivity groups decreases compared to the medium productivity group as the value of career progress increases, which may point towards a curvilinear relationship between the variables. A similar pattern was also found in the model with the citation impact group as the dependent variable, although it was not as apparent.

As a result of this apparent curvilinear relationship, we included quadric terms for career progress in both models, and these were significant. Likelihood ratio tests also show that the models with quadric terms included have a significant better fit to the data. Furthermore, the AIC was also lower for these models compared to the initial models where quadric terms were not included (see Appendix Tables A5 – A7 ). Consequently, we base our results on these models, which can be found in Appendix Table A7 . Due to a low number of respondents in the low categories of the scientific curiosity/discovery variable, we also combined the first three values into one to include it as a variable in the regression analysis, which results in a reduced three-value variable for scientific curiosity/discovery.

Results– productivity percentile group

Using the productivity percentile group as the dependent variable, we find that the motivational aspects of practical application and career progress have a significant effect on the probability of being in the low, medium, or high productivity group but not curiosity/scientific discovery. In Figs.  2 and 3 , each line represents the probability of being in each group across the scale of each motivational aspect.

Predicted probability for being in each of the productivity groups according to the value on the ‘practical application’ variable

Predicted probability of being in the low and high productivity groups according to the value on the ‘progress in my career’ variable

Figure  2 shows that at low values of application, there are no significant differences between the probability of being in either of the groups. However, from around value 3 of application, the differences between the probability of being in each group increases, and these are also significant. As a result, we concluded that high scores on practical application is related to increased probability of being in the high productivity group.

In Fig.  3 , we excluded the medium productivity group from the figure because there are no significant differences between this group and the high and low productivity group. Nevertheless, we found significant differences between the low productivity and the high productivity group. Since we added a quadric term for career progress, the two lines in Fig.  3 have a curvilinear shape. Figure  3 shows that there are only significant differences between the probability of being in the low or high productivity group at mid and high values of career progress. In addition, the probability of being in the high productivity group is at its highest value at mid values of career progress. This indicates that being motivated by career progress increases the probability of being in the high productivity group but only up to a certain point before it begins to have a negative effect on the probability of being in this group.

We also included age and gender as variables in the model, and Figs.  4 and 5 show the results. Figure  4 shows that age especially impacts the probability of being in the high productivity and low productivity groups. The lowest age category (< 30–34 years) has the highest probability for being in the low productivity group, while from the mid age category (50 years and above), the probability is highest for being in the high productivity group. This means that increased age is related to an increased probability of high productivity. The variable controlling for the effect of funding also showed some significant results (see Appendix Table A7 ). The most relevant finding is that receiving competitive grants from external public sources had a very strong and significant positive effect on being in the high productivity group and a medium-sized significant negative effect on being in the low productivity group. This shows that receiving external funding in the form of competitive grants has a strong effect on productivity.

Predicted probability of being in each of the productivity groups according to age

Figure  5 shows that there is a difference between male and female respondents. For females, there are no differences in the probability of being in either of the groups, while males have a higher probability of being in the high productivity group compared to the medium and low productivity groups.

Results– citation impact group

For the citation impact group as the dependent variable, we found that career progress has a significant effect on the probability of being in the low citation impact group or the high citation group but not curiosity/scientific discovery or practical application. Figure  6 shows how the probability of being in the high citation impact group increases as the value on career progress increases and is higher than that of being in the low citation impact group, but only up to a certain point. This indicates that career progress increases the probability of being in the high citation impact group to some degree but that too high values are not beneficial for high citation impact. However, it should also be noted that the effect of career progress is weak and that it is difficult to conclude on how very low or very high values of career progress affect the probability of being in the two groups.

Predicted probability for being in each of the citation impact groups according to the value on the ‘progress in my career’ variable

We also included age and gender as variables in the model, and we found a similar pattern as in the model with productivity percentile group as the dependent variable. However, the relationship between the variables is weaker in this model with the citation impact group as the dependent variable. Figure  7 shows that the probability of being in the high citation impact group increases with age, but there is no significant difference between the probability of being in the high citation impact group and the medium citation impact group. We only see significant differences when each of these groups is compared to the low citation impact group. In addition, the increase in probability is more moderate in this model.

Predicted probability of being in each of the citation impact groups according to age

Figure  8 shows that there are differences between male and female respondents. Male respondents have a significant higher probability of being in the medium or high citation impact group compared to the low citation impact group, but there is no significant difference in the probability between the high and medium citation impact groups. For female respondents, there are no significant differences. Similarly, for age, the effect also seems to be more moderate in this model compared to the model with productivity percentile groups as the dependent variable. In addition, the effect of funding sources is more moderate on citation impact compared to productivity (see Appendix Table A7 ). Competitive grants from external public sources still have the most relevant effect, but the effect size and level of significance is lower than for the model where productivity groups are the dependent variable. Respondents who received a large amount of external funding through competitive grants are more likely to be highly cited, but the effect size is much smaller, and the result is only significant at p  < 0.1. Those who do not receive much funding from this source are more likely to be in the low impact group. Here, the effect size is large, and the coefficient is highly significant.

Predicted probability for being in each of the citation impact groups according to gender

Concluding discussion

This article aimed to explore researchers’ motivations and investigate the impact of motivation on research performance. By addressing these issues across several fields and countries, we provided new evidence on the motivation and performance of researchers.

Most researchers in our large-N survey found curiosity/scientific discovery to be a crucial motivational factor, with practical application being the second most supported aspect. Only a smaller number of respondents saw career progress as an important inspiration to conduct their research. This supports the notion that researchers are mainly motivated by core aspects of academic work such as curiosity, discoveries, and practical application of their knowledge and less so by personal gains (see Evans and Meyer 2003 ). Therefore, our results align with earlier research on motivation. In their interview study of scientists working at a government research institute in the UK, Jindal-Snape and Snape ( 2006 ) found that the scientists were typically motivated by the ability to conduct high quality, curiosity-driven research and de-motivated by the lack of feedback from management, difficulty in collaborating with colleagues, and constant review and change. Salaries, incentive schemes, and prospects for promotion were not considered a motivator for most scientists. Kivistö and colleagues ( 2017 ) also observed similar patterns in more recent survey data from Finnish academics.

As noted in the introduction, the issue of motivation has often been analysed in the literature using the intrinsic-extrinsic distinction. In our study, we have not applied these concepts directly. However, it is clear that the curiosity/scientific discovery item should be considered a type of intrinsic motivation, as it involves performing the activity for its inherent satisfaction. Moreover, the practical application item should probably be considered mainly intrinsic, as it involves creating a better society (for others) without primarily focusing on gains for oneself. The career progress item explicitly mentions personal gains such as position and higher salary and is, therefore, a type of extrinsic motivation. This means that our results support the notion that there are very strong elements of intrinsic motivation among researchers (Jindal-Snape and Snape 2006 ).

When analysing the three aspects of motivation, we found some differences. Physicists tend to view practical application as less important than researchers in the two other fields, while career progress was most emphasised by economists. Regarding country differences, our data suggest that career progress is most important for researchers in Denmark. Nevertheless, given the limited effect sizes, the overall picture is that motivational factors seem to be relatively similar regarding disciplinary and country dimensions.

Regarding gender aspects of motivation, our data show that women seem to view practical application and career progress as more important than men. One explanation for this could be the continued gender differences in academic careers, which tend to disadvantage women, thus creating a greater incentive for female scholars to focus on and be motivated by career progress aspects (Huang et al. 2020 ; Lerchenmueller and Sorenson 2018 ). Unsurprisingly, respondents’ age and academic position influenced the importance of different aspects of motivation, especially regarding career progress. Here, increased age and moving up the positional hierarchy are linked to a decrease in importance. This highlights that older academics and those in more senior positions drew more motivation from other sources that are not directly linked to their personal career gains. This can probably be explained by the academic career ladder plateauing at a certain point in time, as there are often no additional titles and very limited recognition beyond becoming a full professor. Finally, the type of funding that scholars received also had an influence on their productivity and, to a certain extent, citation impact.

Overall, there is little support that researchers across various fields and countries are very different when it comes to their motivation for conducting research. Rather, there seems to be a strong common core of academic motivation that varies mainly by gender and age/position. Rather than talking about researchers’ motivation per se, our study, therefore, suggests that one should talk about motivation across gender, at different stages of the career, and, to a certain degree, in different fields. Thus, motivation seems to be a multi-faceted construct, and the importance of different aspects of motivation vary between different groups.

In the second step of our analysis, we linked motivation to performance. Here, we focused on both scientific productivity and citation impact. Regarding the former, our data show that both practical application and career progress have a significant effect on productivity. The relationship between practical application aspects and productivity is linear, meaning that those who indicate that this aspect of motivation is very important to them have a higher probability of being in the high productivity group. The relationship between career aspects of motivation and productivity is curve linear, and we found only significant differences between the high and low productivity groups at mid and high values of the motivation scale. This indicates that being more motivated by career progress increases productivity but only to a certain extent before it starts having a detrimental effect. A common assumption has been that intrinsic motivation has a positive and instrumental effect and extrinsic motivation has a negative effect on the performance of scientists (Peng and Gao 2019 ; Ryan and Berbegal-Mirabent 2016 ). Our results do not generally support this, as motives related to career progress are positively linked with productivity only to a certain point. Possibly, this can be explained by the fact that the number of publications is often especially important in the context of recruitment and promotion (Langfeldt et al. 2021 ; Reymert et al. 2021 ). Thus, it will be beneficial from a scientific career perspective to have many publications when trying to get hired or promoted.

Regarding citation impact, our analysis highlights that only the career aspects of motivation have a significant effect. Similar to the results regarding productivity, being more motivated by career progress increases the probability of being in the high citation impact group, but only to a certain value when the difference stops being significant. It needs to be pointed out that the effect strength is weaker than in the analysis that focused on productivity. Thus, these results should be treated with greater caution.

Overall, our results shed light on some important aspects regarding the motivation of academics and how this translates into research performance. Regarding our first research question, it seems to be the case that there is not one type of motivation but rather different contextual mixes of motivational aspects that are strongly driven by gender and the academic position/age. We found only limited effects of research fields and even less pronounced country effects, suggesting that while situational, the mix of motivational aspects also has a common academic core that is less influenced by different national environments or disciplinary standards. Regarding our second research question, our results challenge the common assumption that intrinsic motivation has a positive effect and extrinsic motivation has a negative effect on the performance of scientists. Instead, we show that motives related to career are positively linked to productivity at least to a certain point. Our analysis regarding citation patterns achieved similar results. Combined with the finding regarding the importance of current academic position and age for specific patterns of motivation, it could be argued that the fact that the number of publications is often used as a measurement in recruitment and promotion makes academics that are more driven by career aspects publish more, as this is perceived as a necessary condition for success.

Our study has a clear focus on the research side of academic work. However, most academics do both teaching and research, which raises the question of how far our results can also inform our knowledge regarding the motivation for teaching. On the one hand, previous studies have highlighted that intrinsic motivation is also of high importance for the quality of teaching (see e.g. Wilkesmann and Lauer 2020 ), which fits well with our findings. At the same time, the literature also highlights persistent goal conflicts of academics (see e.g. Daumiller et al. 2020 ), given that extra time devoted to teaching often comes at the costs of publications and research. Given that other findings in the literature show that research performance continues to be of higher importance than teaching in academic hiring processes (Reymert et al. 2021 ), the interplay between research performance, teaching performance, and different types of motivation is most likely more complicated and demands further investigation.

While offering several relevant insights, our study still comes with certain limitations that must be considered. First, motivation is a complex construct. Thus, there are many ways one could operationalise it, and not one specific understanding so far seems to have emerged as best practice. Therefore, our approach to operationalisation and measurement should be seen as an addition to this broader field of measurement approaches, and we do not claim that this is the only sensible way of doing it. Second, we rely on self-reported survey data to measure the different aspects of motivation in our study. This means that aspects such as social desirability could influence how far academics claim to be motivated by certain aspects. For example, claiming to be mainly motivated by personal career gains may be considered a dubious motive among academics.

With respect to the bibliometric analyses, it is important to realise that we have lumped researchers into categories, thereby ‘smoothening’ the individual performances into group performances under the various variables. This has an effect that some extraordinary scores might have become invisible in our study, which might have been interesting to analyse separately, throwing light on the relationships we studied. However, breaking the material down to the lower level of analysis of individual researchers also comes with a limitation, namely that at the level of the individual academic, bibliometrics tend to become quite sensitive for the underlying numbers, which in itself is then hampered by the coverage of the database used, the publishing cultures in various countries and fields, and the age and position of the individuals. Therefore, the level of the individual academic has not been analysed in our study, how interesting and promising outcomes might have been. even though we acknowledge that such a study could yield interesting results.

Finally, our sample is drawn from northwestern European countries and a limited set of disciplines. We would argue that we have sufficient variation in countries and disciplines to make the results relevant for a broader audience context. While our results show rather small country or discipline differences, we are aware that there might be country- or discipline-specific effects that we cannot capture due to the sampling approach we used. Moreover, as we had to balance sufficient variation in framework conditions with the comparability of cases, the geographical generalisation of our results has limitations.

This article investigated what motivates researchers across different research fields and countries and how this motivation influences their research performance. The analysis showed that the researchers are mainly motivated by scientific curiosity and practical application and less so by career considerations. Furthermore, the analysis shows that researchers driven by practical application aspects of motivation have a higher probability of high productivity. Being driven by career considerations also increases productivity but only to a certain extent before it starts having a detrimental effect.

The article is based on a large-N survey of economists, cardiologists, and physicists in Denmark, Norway, Sweden, the Netherlands, and the UK. Building on this study, future research should expand the scope and study the relationship between motivation and productivity as well as citation impact in a broader disciplinary and geographical context. In addition, we encourage studies that develop and validate our measurement and operationalisation of aspects of researchers’ motivation.

Finally, a long-term panel study design that follows respondents throughout their academic careers and investigates how far their motivational patterns shift over time would allow for more fine-grained analysis and thereby a richer understanding of the important relationship between motivation and performance in academia.

Data availability

The data set for this study is available from the corresponding author upon reasonable request.

Aksnes DW, Sivertsen G (2019) A criteria-based assessment of the coverage of Scopus and web of Science. J Data Inform Sci 4(1):1–21. https://doi.org/10.2478/jdis-2019-0001

Article   Google Scholar  

Atta-Owusu K, Fitjar RD (2021) What motivates academics for external engagement? Exploring the effects of motivational drivers and organizational fairness. Sci Public Policy. https://doi.org/10.1093/scipol/scab075 . November, scab075

Baccini A, Barabesi L, Cioni M, Pisani C (2014) Crossing the hurdle: the determinants of individual. Sci Perform Scientometrics 101(3):2035–2062. https://doi.org/10.1007/s11192-014-1395-3

Bornmann L, Leydesdorff L, Mutz R (2013) The use of percentiles and percentile rank classes in the analysis of bibliometric data: opportunities and limits. J Informetrics 7(1):158–165. https://doi.org/10.1016/j.joi.2012.10.001

Cruz-Castro L, Sanz-Menendez L (2021) What should be rewarded? Gender and evaluation criteria for tenure and promotion. J Informetrics 15(3):1–22. https://doi.org/10.1016/j.joi.2021.101196

Daumiller M, Stupnisky R, Janke S (2020) Motivation of higher education faculty: theoretical approaches, empirical evidence, and future directions. Int J Educational Res 99:101502. https://doi.org/10.1016/j.ijer.2019.101502

Duarte H, Lopes D (2018) Career stages and occupations impacts on workers motivations. Int J Manpow 39(5):746–763. https://doi.org/10.1108/IJM-02-2017-0026

Evans IM, Meyer LH (2003) Motivating the professoriate: why sticks and carrots are only for donkeys. High Educ Manage Policy 15(3):151–167. https://doi.org/10.1787/hemp-v15-art29-en

Finkelstein MJ (1984) The American academic profession: a synthesis of social scientific inquiry since World War II. Ohio State University, Columbus

Google Scholar  

Hammarfelt B, de Rijcke S (2015) Accountability in context: effects of research evaluation systems on publication practices, disciplinary norms, and individual working routines in the Faculty of arts at Uppsala University. Res Evaluation 24(1):63–77. https://doi.org/10.1093/reseval/rvu029

Hangel N, Schmidt-Pfister D (2017) Why do you publish? On the tensions between generating scientific knowledge and publication pressure. Aslib J Inform Manage 69(5):529–544. https://doi.org/10.1108/AJIM-01-2017-0019

Hazelkorn E (2015) Rankings and the reshaping of higher education: the battle for world-class excellence. Palgrave McMillan, Basingstoke

Book   Google Scholar  

Hilbe JM (2017) Logistic regression models. Taylor & Francis Ltd, London

Horodnic IA, Zaiţ A (2015) Motivation and research productivity in a university system undergoing transition. Res Evaluation 24(3):282–292

Huang J, Gates AJ, Sinatra R, Barabási A-L (2020) Historical comparison of gender inequality in scientific careers across countries and disciplines. Proceedings of the National Academy of Sciences 117(9):4609–4616. https://doi.org/10.1073/pnas.1914221117

Jeong S, Choi JY, Kim J-Y (2014) On the drivers of international collaboration: the impact of informal communication, motivation, and research resources. Sci Public Policy 41(4):520–531. https://doi.org/10.1093/scipol/sct079

Jindal-Snape D, Snape JB (2006) Motivation of scientists in a government research institute: scientists’ perceptions and the role of management. Manag Decis 44(10):1325–1343. https://doi.org/10.1108/00251740610715678

Kivistö J, Pekkola E, Lyytinen A (2017) The influence of performance-based management on teaching and research performance of Finnish senior academics. Tert Educ Manag 23(3):260–275. https://doi.org/10.1080/13583883.2017.1328529

Kulczycki E, Engels TCE, Pölönen J, Bruun K, Dušková M, Guns R et al (2018) Publication patterns in the social sciences and humanities: evidence from eight European countries. Scientometrics 116(1):463–486. https://doi.org/10.1007/s11192-018-2711-0

Lam A (2011) What motivates academic scientists to engage in research commercialization: gold, ribbon or puzzle? Res Policy 40(10):1354–1368. https://doi.org/10.1016/j.respol.2011.09.002

Langfeldt L, Reymert I, Aksnes DW (2021) The role of metrics in peer assessments. Res Evaluation 30(1):112–126. https://doi.org/10.1093/reseval/rvaa032

Larivière V, Macaluso B, Archambault É, Gingras Y (2010) Which scientific elites? On the concentration of research funds, publications and citations. Res Evaluation 19(1):45–53. https://doi.org/10.3152/095820210X492495

Lepori B, Jongbloed B, Hicks D (2023) Introduction to the handbook of public funding of research: understanding vertical and horizontal complexities. In: Lepori B, Hicks BJ D (eds) Handbook of public funding of research. Edward Elgar Publishing, Cheltenham, pp 1–19

Chapter   Google Scholar  

Lerchenmueller MJ, Sorenson O (2018) The gender gap in early career transitions in the life sciences. Res Policy 47(6):1007–1017. https://doi.org/10.1016/j.respol.2018.02.009

Leslie DW (2002) Resolving the dispute: teaching is academe’s core value. J High Educ 73(1):49–73

Lounsbury JW, Foster N, Patel H, Carmody P, Gibson LW, Stairs DR (2012) An investigation of the personality traits of scientists versus nonscientists and their relationship with career satisfaction: relationship of personality traits and career satisfaction of scientists and nonscientists. R&D Manage 42(1):47–59. https://doi.org/10.1111/j.1467-9310.2011.00665.x

Ma L (2019) Money, morale, and motivation: a study of the output-based research support scheme. Univ Coll Dublin Res Evaluation 28(4):304–312. https://doi.org/10.1093/reseval/rvz017

Melguizo T, Strober MH (2007) Faculty salaries and the maximization of prestige. Res High Educt 48(6):633–668

Moed HF (2005) Citation analysis in research evaluation. Springer, Dordrecht

Netherlands Observatory of Science (NOWT) (2012) Report to the Dutch Ministry of Science, Education and Culture (OC&W). Den Haag 1998

Peng J-E, Gao XA (2019) Understanding TEFL academics’ research motivation and its relations with research productivity. SAGE Open 9(3):215824401986629. https://doi.org/10.1177/2158244019866295

Piro FN, Aksnes DW, Rørstad K (2013) A macro analysis of productivity differences across fields: challenges in the measurement of scientific publishing. J Am Soc Inform Sci Technol 64(2):307–320. https://doi.org/10.1002/asi.22746

Pruvot EB, Estermann T, Popkhadze N (2023) University autonomy in Europe IV. The scorecard 2023. Retrieved from Brussels. https://eua.eu/downloads/publications/eua autonomy scorecard.pdf

Reymert I, Jungblut J, Borlaug SB (2021) Are evaluative cultures national or global? A cross-national study on evaluative cultures in academic recruitment processes in Europe. High Educ 82(5):823–843. https://doi.org/10.1007/s10734-020-00659-3

Roach M, Sauermann H (2010) A taste for science? PhD scientists’ academic orientation and self-selection into research careers in industry. Res Policy 39(3):422–434. https://doi.org/10.1016/j.respol.2010.01.004

Rørstad K, Aksnes DW (2015) Publication rate expressed by age, gender and academic position– A large-scale analysis of Norwegian academic staff. J Informetrics 9(2):317–333. https://doi.org/10.1016/j.joi.2015.02.003

Ruiz-Castillo J, Costas R (2014) The skewness of scientific productivity. J Informetrics 8(4):917–934. https://doi.org/10.1016/j.joi.2014.09.006

Ryan JC (2014) The work motivation of research scientists and its effect on research performance: work motivation of research scientists. R&D Manage 44(4):355–369. https://doi.org/10.1111/radm.12063

Ryan JC, Berbegal-Mirabent J (2016) Motivational recipes and research performance: a fuzzy set analysis of the motivational profile of high-performing research scientists. J Bus Res 69(11):5299–5304. https://doi.org/10.1016/j.jbusres.2016.04.128

Ryan RM, Deci EL (2000) Intrinsic and extrinsic motivations: classic definitions and new directions. Contemp Educ Psychol 25(1):54–67. https://doi.org/10.1006/ceps.1999.1020

Sivertsen G (2019) Understanding and evaluating research and scholarly publishing in the social sciences and humanities (SSH). Data Inform Manage 3(2):61–71. https://doi.org/10.2478/dim-2019-0008

Sivertsen G, Van Leeuwen T (2014) Scholarly publication patterns in the social sciences and humanities and their relationship with research assessment

Stephan P, Veugelers R, Wang J (2017) Reviewers are blinkered by bibliometrics. Nature 544(7651):411–412. https://doi.org/10.1038/544411a

Thomas D, Nedeva M (2012) Characterizing researchers to study research funding agency impacts: the case of the European Research Council’s starting grants. Res Evaluation 21(4):257–269. https://doi.org/10.1093/reseval/rvs020

Tien FF (2000) To what degree does the desire for promotion motivate faculty to perform research? Testing the expectancy theory. Res High Educt 41(6):723–752. https://doi.org/10.1023/A:1007020721531

Tien FF (2008) What kind of faculty are motivated to perform research by the desire for promotion? High Educ 55(1):17–32. https://doi.org/10.1007/s10734-006-9033-5

Tien FF, Blackburn RT (1996) Faculty rank system, research motivation, and faculty research productivity: measure refinement and theory testing. J High Educ 67(1):2. https://doi.org/10.2307/2943901

Vallerand RJ, Pelletier LG, Blais MR, Briere NM, Senecal C, Vallieres EF (1992) The academic motivation scale: a measure of intrinsic, extrinsic, and amotivation in education. Educ Psychol Meas 52(4):1003–1017. https://doi.org/10.1177/0013164492052004025

Van Iddekinge CH, Aguinis H, Mackey JD, DeOrtentiis PS (2018) A meta-analysis of the interactive, additive, and relative effects of cognitive ability and motivation on performance. J Manag 44(1):249–279. https://doi.org/10.1177/0149206317702220

Van Leeuwen T (2013) Bibliometric research evaluations, Web of Science and the social sciences and humanities: A problematic relationship? Bibliometrie - Praxis Und Forschung, September, Bd. 2(2013). https://doi.org/10.5283/BPF.173

Van Leeuwen T, van Wijk E, Wouters PF (2016) Bibliometric analysis of output and impact based on CRIS data: a case study on the registered output of a Dutch university. Scientometrics 106(1):1–16. https://doi.org/10.1007/s11192-015-1788-y

Waltman L, Schreiber M (2013) On the calculation of percentile-based bibliometric indicators. J Am Soc Inform Sci Technol 64(2):372–379. https://doi.org/10.1002/asi.22775

Waltman L, van Eck NJ, van Leeuwen TN, Visser MS, van Raan AFJ (2011) Towards a new crown indicator: an empirical analysis. Scientometrics 87(3):467–481. https://doi.org/10.1007/s11192-011-0354-5

Wilkesmann U, Lauer S (2020) The influence of teaching motivation and new public management on academic teaching. Stud High Educ 45(2):434–451. https://doi.org/10.1080/03075079.2018.1539960

Wilsdon J, Allen L, Belfiore E, Campbell P, Curry S, Hill S, Jones R et al (2015) The metric tide: report of the independent review of the role of metrics in research assessment and management. https://doi.org/10.13140/RG.2.1.4929.1363

Zacharewicz T, Lepori B, Reale E, Jonkers K (2019) Performance-based research funding in EU member states—A comparative assessment. Sci Public Policy 46(1):105–115. https://doi.org/10.1093/scipol/scy041

Zhang L, Sivertsen G, Du H, Huang Y, Glänzel W (2021) Gender differences in the aims and impacts of research. Scientometrics 126(11):8861–8886. https://doi.org/10.1007/s11192-021-04171-y

Download references

Acknowledgements

We are thankful to the R-QUEST team for input and comments to the paper.

The authors disclosed the receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Research Council Norway (RCN) [grant number 256223] (R-QUEST).

Open access funding provided by University of Oslo (incl Oslo University Hospital)

Author information

Silje Marie Svartefoss

Present address: TIK Centre for Technology, Innovation and Culture, University of Oslo, 0317, Oslo, Norway

Authors and Affiliations

Nordic Institute for Studies in Innovation, Research and Education (NIFU), Økernveien 9, 0608, Oslo, Norway

Silje Marie Svartefoss & Dag W. Aksnes

Department of Political Science, University of Oslo, 0315, Oslo, Norway

Jens Jungblut & Kristoffer Kolltveit

Centre for Science and Technology Studies (CWTS), Leiden University, 2311, Leiden, The Netherlands

Thed van Leeuwen

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Silje Marie Svartefoss, Jens Jungblut, Dag W. Aksnes, Kristoffer Kolltveit, and Thed van Leeuwen. The first draft of the manuscript was written by all authors in collaboration, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Silje Marie Svartefoss .

Ethics declarations

Competing interests.

The authors have no competing interests to declare that are relevant to the content of this article.

Informed consent

was retrieved from the participants in this study.

Electronic Supplementary Material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Svartefoss, S.M., Jungblut, J., Aksnes, D.W. et al. Explaining research performance: investigating the importance of motivation. SN Soc Sci 4 , 105 (2024). https://doi.org/10.1007/s43545-024-00895-9

Download citation

Received : 14 December 2023

Accepted : 15 April 2024

Published : 23 May 2024

DOI : https://doi.org/10.1007/s43545-024-00895-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Performance
• Productivity
• Find a journal
• Publish with us
• Track your research

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Transl Med UniSa
• v.5; Jan-Apr 2013

Statistics and Medicine: the Indispensable Know-How of the Researcher

1 Department of Medicine, University of Salerno

2 Department of UTTP-FOS, ENEA

Statistics has often been misunderstood in Medicine, but it is indispensable knowledge both for the experimenter and the reader. Statistical methods allow to study diseases, patients, and epidemiological events. The modern researcher cannot refuse to know and to use statistics. A deeper understanding is required to prepare a research project and to avoid colossal mistakes of misleading.

The aim of this paper is to provide an organized and structured point of view on the use of statistics in Medicine and Research, showing the principal resources to organize a scientific study from the declaration of a hypothesis to the report of the results.

“Statistical thinking will one day be as necessary for efficient citizenship as the ability to read or write.” H. G. Wells

I. INTRODUCTION

Statistics represents the heel of Achilles for the modern researcher. The necessity to apply statistics to any scientific problem leads the researcher to assume the burden of a deeper knowledge in this science.

Medicine often uses probabilistic statistics that could be far away from the scientific method. Data are used and analysed in order to highlight trends or to make a prevision for the validity of a diagnostic method, a therapy or a prognosis for a disease. When an event is pointed out in a large amount of cases, a descriptive method is used to identify the potential basic mechanism. Starting from this point, the aim of the researcher should be the precise and reasoned planning of a scientific project. Any aspect of a research project should be provided at the beginning and then verified.

The data from the study of a subset of the population, the sample, are collected and then subjected to statistical analysis, in order to attend to two purposes: description and inference.

Descriptive statistics uses data to describe numerically and graphically the observations in the sample. Instead, inferential statistics uses the identification of patterns in data to draw inferences about the population.

Statistical analysis might identify the presence of a relationship between variables.

The need for greater knowledge in statistics is crucial both for the researcher, who is planning a study or elaborating data from a study, and for readers to understand the plan of a study and the validity of its conclusion.

II. STATISTICAL METHODS

A) experimental and observational studies.

The use of statistical studies allows to evaluate the causality and to draw a conclusion about the effect of an independent variable on the dependent variables examined. We consider two types of statistical study: experimental and observational ( Table 1 ). They differ in how studies are conducted and in their aims [ 1 , 2 ].

Classification of observational and experimental studies.

An observational study evaluates the inferences about the effect of a treatment on patients, when the option of giving a patient to a treated group versus a control group is not an option for the investigator. Data are already available, researchers can explore any similarities among them. So, it is characterized by the observation of the interaction between two or more variables: the independent variable (or the risk factor) and, on the other hand, the dependent variables (or the response). This type of study might be prospective (with the evaluation of the relative risk) or retro-prospective (with the evaluation of the odds ratio). It is essential to consider the presence of confounding variables that influenced the relation between the risk factor and the response causing a bias. This study might be preliminary to an experimental study, the starting point to plan a research.

The experimental model is fundamental for the scientific model: after the observation of a phenomenon, the hypothesis is declared and then verified. Nowadays, planning a research project and identifying a hypothesis clearly, it is necessary to check its innovation. Only after that the researcher can design and perform the experiment and at the end evaluate the validity of the hypothesis.

The substantial difference between the observational and experimental study is in the role of the researcher. In an experimental study researchers are involved in taking measurements, manipulating the system, and then taking again measurements using the same procedure in order to understand the impact of the manipulation on the system. Instead, in an observational study there is no experimental manipulation but only the data collection and their analysis.

Designing an experimental study, the researcher needs to:

• consider the initial information about the entity of the effects, the alternative hypothesis and the experimental variability;
• plan the research and design experiments, choosing the quantity to measure, the independent and the dependent variables, the method of measurement, the acceptable error of measure, the variability range, the acceptable influence of confounding variables;
• perform the experiment and measurement;
• analyse data following strict determined before;
• present the results of the study.

Planning a study the researcher decides to support an “hypothesis”, a conjecture about the variables taken into account in a population. This hypothesis can be verified with statistical method, e.g. confirmatory tests and χ 2 test.

B) The choice between parametric and non-parametric methods

A parametric model is a distribution that can be described using a finite number of parameters. Many basic statistical methods are parametric. The use of parametric methods leads to make more assumptions than using non-parametric methods. If these assumptions are correct, parametric methods can be accurate and precise.

The necessity to make assumptions is a risk, because if they are wrong, the parametric method will be misleading. This is the reason why they are considered statistical powerful, but not robust.

In addition to parametric models, there are also semi-parametric, semi-nonparametric, and non-parametric models. A non-parametric model is a model where all parameters are in infinite-dimensional parameter spaces. It considers data belonging to any particular distribution or techniques.

The use of non-parametric methods may be indispensable when data are characterized by a ranking without an useful numerical interpretation. Non-parametric methods make fewer assumptions than parametric methods. So they can be applied widely and be considered robust. Nonparametric tests have less power where a parametric test would be appropriate [ 3 ].

C) The evaluation and control of the error

The choice of a sample might be extremely difficult. The sample should be representative of an entire population to be used as a guide. The use of a representative sample assures that the inferences and conclusions can be safely extended from the sample to all the population. A crucial problem is in determining the adequate size of the sample. The sample should be large enough to be representative, but it does not need to be oversize to reduce the costs of the study.

There are two large groups of errors:

• the error due to measurement instrumentation and to the operator (for example, systematic error, sensitivity error). It is a controllable but often unavoidable error, due to a discretional evaluation of the measurement. In the medical field, much emphasis is given to error “bias” is an error of this kind due to experimenter’s choices.
• the statistical error, that is always present when there is a measurement, has many causes, even inherent to the experiment itself. This error can be reduced increasing the number of measurements, but it cannot be eliminated. It is the measure of the amplitude of data sheet dispersion. The normal distribution is frequently used to describe random variables to real value. It is considered the most prominent distribution in statistics because it is a distribution that can be used for a large number of random variables. It is commonly used in practice as a simple model for complex phenomena both in natural science and in social science. The error in an experiment is usually assumed to follow a normal distribution. This assumption is used to calculate the propagation of uncertainty. When the distribution is a Gaussian distribution the error α can be calculate as (1) Δ x = ∑ ( x i − μ ) 2 ( n − 1 ) (1) where n is the sample size and μ is the mean of the distribution [ 3 ].

Many of the errors derived from the measurements are classified as “noise” if they are random. They can lead to statistic misleading. Misuse of statistics can produce subtle, but serious errors in description and interpretation. Even experienced professionals make such errors, but they can lead to enormous decision errors.

D) Significance levels

Statistics offers some instruments to understand if the results of the measurements is that expected by the hypothesis. One of these is the significance test.

A significance test evaluates the plausibility of the observed data when a “null hypothesis” is true. It is expressed by a significance level α that is the probability of rejecting a null hypothesis that is true [ 3 ].

In the interpretation of the statistical information, it is possible to evaluate the null hypothesis wrongly.

As reported in Table 2 , there are two types of error:

• Type I error or α, when the null hypothesis is rejected, but it is true. It is a “false positive”.
• Type II error or β, when the null hypothesis is false, but it is not rejected. It is a “false negative”.

The null hypothesis and the possible errors.

Rejecting a true null hypothesis is an error that should be as small as possible. It is necessary to choose the smallest value: 0.01, 0.05 or 0.001. This significance level is named P value. It is used to express the chance that the observed data are due to chance. The relevance of the chosen P value might depend on the sample size. Furthermore, reporting results it is important to report the real P value.

It is difficult to choose a significance level, but it is extremely difficult interpreting results using the P value. If P value of observed data is small enough (P< 0.05 or P< 0.01 or P< 0.001), the null hypothesis could be rejected, and the alternative hypothesis could be accepted. But if P value is not small enough, there is no statistical significance. The null hypothesis cannot be rejected, but it does not mean anything else.

E) Survival statistics

Survival statistics is a branch of statistics that deals with death in a biological organism. Time is a fundamental character of this branch and the main “events” are death and failure. Survival models can be usefully considered as ordinary regression models in which time is the response variable.

Among the estimators used to estimating survival data, the Kaplan–Meier estimator is extremely important. In the medical literature, it is frequently used to measure the rate of patients living after a therapy or after the exposition to a risk factor. A plot of the Kaplan–Meier estimate of patients’ survival is made by horizontal steps of declining magnitude. The value of the survival function is constant between clicks that are successive distinct sampled observations.

Among survival analysis, it is necessary to consider the survival rate. It is percentage of people alive after 1, 5 or 10 years after the diagnosis of disease. It can be expressed as relative survival, cause specific survival, disease specific survival. The interval considered can be one, five, and ten years. Survival rates can be used to compare the effectiveness of treatments and to evaluated the prognosis of patients [ 1 ].

III. REPORTING STATISTICAL RESULTS

Planning and reporting an observational or an experimental study it is important to declare the aim of the study, the statistical method that will be used and the endpoints. The characteristics of subjects and of the planning should be clearly expressed. It is important if the study is blinding (masking). The endpoints are the events that will be considered. It is possible to differentiate primary and secondary endpoints. Examples of endpoints can be death (in a study survival rate), adverse events and morbidity, toxicity, maximum tolerated dose, response to a therapy. Every adverse event should be reported.

The software used for analysing data should be reported, e.g. SPSS or SAS. So, statistical tests should be described. There should be the type of statistical test, if it is a confirmatory analysis or if it is an exploratory test that need to be strengthen with cross validation [ 1 , 4 ].

Criteria of inclusion or exclusion are indispensable to clarify the validity of interpretation of the data and the extent to which the results can be applied.

Data should be enlighten and so their sources [ 4 ].

Reporting results, any exclusion during analysis should be expressed. Data analysis should be reported with appropriate confidence intervals or P value [ 3 , 4 ].

The interpretation of the P value is disputable. The presence of a P value less than a predefined value determines statistical significance, but its absence is not a reason to reject the alternative hypothesis. Statistical significance does not mean equivalence of treatment or techniques because the clinical importance of a result depends on several variables. Careful statisticians distinguish between statistical and medical significance.

To avoid problems in the interpretation of analysis, it is necessary to choose easy endpoints, to avoid using the term significant, to use correction for data when it is due, e.g. the correction of Bonferroni.

Results and data should be reported in a clear and not confused way, using tables and graphs. The estimation of a valuable and important relation among data is facilitated using graphs.

Graphs give a visual perception of results and catch attention quickly. Understanding of the context is immediate if a graph is well-organized. A graph has two component: content and format. The content is what is reported in a graph while the format is the way that it is reported. It is fundamental to avoid the overlapping of data symbols. If the graph requires the use of many lines, choose patterns that are distinguished easily from each other [ 5 ].

If visualizing distributions is essential, show the distribution of data with scatter plots or one-way plots, but avoid graphs showing summaries of data such as bar graphs showing means with standard deviation. Good work is a work well reported.

Improving statistical methods to protect wildlife populations

Threats affecting more and more species

In human populations, it is relatively easy to calculate demographic trends and make projections of the future if data on basic processes such as births and immigration is known. The data, given by individuals, can be also death and emigration, which subtract. In the wild, on the other hand, understanding the processes that determine wildlife demographic patterns is a highly complex challenge for the scientific community. Although a wide range of methods are now available to estimate births and deaths in wildlife, quantifying emigration and immigration has historically been difficult or impossible in many populations of interest, particularly in the case of threatened species.

A paper published in the journal Biological Conservation warns that missing data on emigration and immigration movements in wildlife can lead to significant biases in species' demographic projections. As a result, projections about the short-, medium- and long-term future of study populations may be inadequate. This puts their survival at risk due to the implementation of erroneous or ineffective conservation strategies. The authors of the new study are Joan Real, Jaume A. Badia-Boher and Antonio Hernández-Matías, from the Conservation Biology team of the Faculty of Biology of the University of Barcelona and the Institute for Research on Biodiversity (IRBio).

More reliable population predictions

This new study on population biology is based on data collected from 2008 to 2020 on the population of the Bonelli's eagle ( Aquila fasciata ), a threatened species that can be found in Catalonia in the coastal and pre-coastal mountain ranges, from the Empordà to Terres de L'Ebre. In the study, the team emphasises the precision of the population viability analysis (PVA) methodology to improve the management and conservation of long-lived species in the natural environment.

"Population viability analyses are a set of methods that allow us to project the demography of a species into the future, mainly to quantify the probability of extinction of a given species or population of interest," says Joan Real, professor at the Department of Evolutionary Biology, Ecology and Environmental Sciences and head of the Conservation Biology team.

"To date -- he continues -- these projections have mostly been carried out only with data on births and deaths, so that migration processes were ignored because of the difficulty of obtaining these data. In other words, we are trying to make demographic projections without considering two key demographic processes."

In the study of wildlife, population models that do not incorporate immigration or emigration "have a considerable probability of leading to biased projections of future population trends. However, explicitly considering migratory processes allows us to consider all the key demographic processes that determine the future trend of a population," says expert Jaume A. Badia-Boher, first author of the study. "This allows us to be much more precise when making demographic predictions, and therefore also when planning future conservation strategies," he adds.

The development of new and more sophisticated statistical methods over the last decade has made it possible to estimate emigration and immigration in a much more accessible way than before. Including these processes in population viability analyses is therefore relatively straightforward, the paper details.

"This new perspective may imply a relevant advance in the reliability of population viability analyses, which will allow us to estimate the future trend of populations more accurately and propose conservation actions more efficiently," notes Professor Antonio Hernández-Matías. "This is of great importance given that in the current context of global change the extinction rates of species are increasing, and more and more species require urgent and effective conservation actions to reverse their decline," the expert says.

Applying methodological developments to conserve biodiversity

Introducing changes in the structure and modelling of population viability analyses can lead to multiple benefits in many areas of biodiversity research and conservation. "Methodological advances are effective when they are applied. For this reason, the application of the new methodology in populations and species of conservation interest should be promoted. It is a priority to make these methodologies known to the scientific community, managers and administration, in order to prioritise conservation actions with the best available methods," say the authors.

"In the future, new methodologies must continue to be developed, as has been done in this study, as they are key to understanding how wild populations function, what measures need to be implemented to conserve them, and how to make these measures as efficient as possible. In the case of endangered species such as the Bonelli's eagle, knowing the rates of emigration and immigration is key to understanding the state of self-sustainability of a population, and thus implementing efficient conservation measures," concludes the team.

• Ecology Research
• Wild Animals
• Biodiversity
• Environmental Policy
• Population dynamics of fisheries
• Weather forecasting
• Wildlife gardening
• Data mining
• Scientific visualization
• Temperature record of the past 1000 years
• Scientific method
• IPCC Report on Climate Change - 2007

Story Source:

Materials provided by University of Barcelona . Note: Content may be edited for style and length.

Journal Reference :

• Jaume A. Badia-Boher, Joan Real, Antonio Hernández-Matías. Assumptions about survival estimates and dispersal processes can have severe impacts on population viability assessments . Biological Conservation , 2024; 292: 110550 DOI: 10.1016/j.biocon.2024.110550

Cite This Page :

Explore More

• The Case of the Missing Black Holes
• Adjusting Sunglasses for Your Windows
• Novel Gene-Editing Tool Created
• How Hummingbirds Hover With Such Accuracy
• Complete X and Y Chromosomes of Great Apes
• Moonlets Stuck Together Orbit 'Dinky' Asteroid
• Orchids Aid Seedlings Through Fungal Networks
• Precise Maps of the Moon's Surface
• Amazing Expertise of Scent Detection Dogs
• Getting to Grips With a Handy Extra Thumb

Trending Topics

Strange & offbeat.

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• Publish with us
• About the journal
• Meet the editors
• Specialist reviews
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Volume 3, Issue 1
• Regular use of fish oil supplements and course of cardiovascular diseases: prospective cohort study
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• Ge Chen 1 ,
• Zhengmin (Min) Qian 2 ,
• Junguo Zhang 1 ,
• Shiyu Zhang 1 ,
• http://orcid.org/0000-0002-7003-6565 Zilong Zhang 1 ,
• Michael G Vaughn 3 ,
• Hannah E Aaron 2 ,
• Chuangshi Wang 4 ,
• Gregory YH Lip 5 , 6 and
• http://orcid.org/0000-0002-3643-9408 Hualiang Lin 1
• 1 Department of Epidemiology , Sun Yat-Sen University , Guangzhou , China
• 2 Department of Epidemiology and Biostatistics, College for Public Health and Social Justice , Saint Louis University , Saint Louis , Missouri , USA
• 3 School of Social Work, College for Public Health and Social Justice , Saint Louis University , Saint Louis , Missouri , USA
• 4 Medical Research and Biometrics Centre , Fuwai Hospital, National Centre for Cardiovascular Diseases, Peking Union Medical College , Beijing , China
• 5 Liverpool Centre for Cardiovascular Science , University of Liverpool and Liverpool Heart and Chest Hospital , Liverpool , UK
• 6 Department of Clinical Medicine , Aalborg University , Aalborg , Denmark
• Correspondence to Dr Hualiang Lin, Department of Epidemiology, Sun Yat-Sen University, Guangzhou, Guangdong 510080, China; linhualiang{at}mail.sysu.edu.cn

Objective To examine the effects of fish oil supplements on the clinical course of cardiovascular disease, from a healthy state to atrial fibrillation, major adverse cardiovascular events, and subsequently death.

Design Prospective cohort study.

Setting UK Biobank study, 1 January 2006 to 31 December 2010, with follow-up to 31 March 2021 (median follow-up 11.9 years).

Participants 415 737 participants, aged 40-69 years, enrolled in the UK Biobank study.

Main outcome measures Incident cases of atrial fibrillation, major adverse cardiovascular events, and death, identified by linkage to hospital inpatient records and death registries. Role of fish oil supplements in different progressive stages of cardiovascular diseases, from healthy status (primary stage), to atrial fibrillation (secondary stage), major adverse cardiovascular events (tertiary stage), and death (end stage).

Results Among 415 737 participants free of cardiovascular diseases, 18 367 patients with incident atrial fibrillation, 22 636 with major adverse cardiovascular events, and 22 140 deaths during follow-up were identified. Regular use of fish oil supplements had different roles in the transitions from healthy status to atrial fibrillation, to major adverse cardiovascular events, and then to death. For people without cardiovascular disease, hazard ratios were 1.13 (95% confidence interval 1.10 to 1.17) for the transition from healthy status to atrial fibrillation and 1.05 (1.00 to 1.11) from healthy status to stroke. For participants with a diagnosis of a known cardiovascular disease, regular use of fish oil supplements was beneficial for transitions from atrial fibrillation to major adverse cardiovascular events (hazard ratio 0.92, 0.87 to 0.98), atrial fibrillation to myocardial infarction (0.85, 0.76 to 0.96), and heart failure to death (0.91, 0.84 to 0.99).

Conclusions Regular use of fish oil supplements might be a risk factor for atrial fibrillation and stroke among the general population but could be beneficial for progression of cardiovascular disease from atrial fibrillation to major adverse cardiovascular events, and from atrial fibrillation to death. Further studies are needed to determine the precise mechanisms for the development and prognosis of cardiovascular disease events with regular use of fish oil supplements.

• Health policy
• Nutritional sciences
• Public health

Data availability statement

Data are available upon reasonable request. UK Biobank is an open access resource. Bona fide researchers can apply to use the UK Biobank dataset by registering and applying at http://ukbiobank.ac.uk/register-apply/ .

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See:  https://creativecommons.org/licenses/by/4.0/ .

https://doi.org/10.1136/bmjmed-2022-000451

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

Findings of the effects of omega 3 fatty acids or fish oil on the risk of cardiovascular disease are controversial

Most previous studies focused on one health outcome and did not characterise specific cardiovascular disease outcomes (eg, atrial fibrillation, myocardial infarction, stroke, heart failure, and major adverse cardiovascular events)

Whether fish oil could differentially affect the dynamic course of cardiovascular diseases, from atrial fibrillation to major adverse cardiovascular events, to other specific cardiovascular disease outcomes, or even to death, is unclear

WHAT THIS STUDY ADDS

In people with no known cardiovascular disease, regular use of fish oil supplements was associated with an increased relative risk of atrial fibrillation and stroke

In people with known cardiovascular disease, the beneficial effects of fish oil supplements were seen on transitions from atrial fibrillation to major adverse cardiovascular events, atrial fibrillation to myocardial infarction, and heart failure to death

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE, OR POLICY

Regular use of fish oil supplements might have different roles in the progression of cardiovascular disease

Further studies are needed to determine the precise mechanisms for the development and prognosis of cardiovascular disease events with regular use of fish oil supplements

Introduction

Cardiovascular disease is the leading cause of death worldwide, accounting for about one sixth of overall mortality in the UK. 1 2 Fish oil, a rich source of omega 3 fatty acids, containing eicosapentaenoic acid and docosahexaenoic acid, has been recommended as a dietary measure to prevent cardiovascular disease. 3 The UK National Institute for Health and Care Excellence recommends that people with or at high risk of cardiovascular disease consume at least one portion of oily fish a week, and the use of fish oil supplements has become popular in the UK and other western countries in recent years. 4 5

Although some epidemiological and clinical studies have assessed the effect of omega 3 fatty acids or fish oil on cardiovascular disease and its risk factors, the findings are controversial. The Agency for Healthcare Research and Quality systematically reviewed 37 observational studies and 61 randomised controlled trials, and found evidence indicating the beneficial effects of higher consumption of fish oil supplements on ischaemic stroke, whereas no beneficial effect was found for atrial fibrillation, major adverse cardiovascular events, myocardial infarction, total stroke, or all cause death. 6 In contrast, the Reduction of Cardiovascular Events with Icosapent Ethyl-Intervention Trial (REDUCE-IT) reported a decreased risk of major adverse cardiovascular events with icosapent ethyl in patients with raised levels of triglycerides, regardless of the use of statins. 7 Most of these findings, however, tended to assess the role of fish oil at a certain stage of cardiovascular disease. For example, some studies restricted the study population to people with a specific cardiovascular disease or at a high risk of cardiovascular disease, 8 9 whereas others evaluated databases of generally healthy populations. 10 All of these factors might preclude direct comparison of the effects of omega 3 fatty acids on atrial fibrillation events or on further deterioration of cardiovascular disease. Few studies have fully characterised specific cardiovascular disease outcomes or accounted for differential effects based on the complex disease characteristics of participants. Hence, in this study, we hypothesised that fish oil supplements might have harmful, beneficial, or no effect on different cardiovascular disease events in patients with varying health conditions.

Most previous studies on the association between fish oil and cardiovascular diseases generally focused on one health outcome. Also, no study highlighted the dynamic progressive course of cardiovascular diseases, from healthy status (primary stage), to atrial fibrillation (secondary stage), major adverse cardiovascular events (tertiary stage), and death (end stage). Clarifying this complex pathway in relation to the detailed progression of cardiovascular diseases would provide substantial insights into the prevention or treatment of future disease at critical stages. Whether fish oil could differentially affect the dynamic course of cardiovascular disease (ie, from atrial fibrillation to major adverse cardiovascular events, to other specific cardiovascular disease outcomes, or even to death) is unclear.

To deal with this evidence gap, we conducted a longitudinal cohort study to estimate the associations between fish oil supplements and specific clinical cardiovascular disease outcomes, including atrial fibrillation, major adverse cardiovascular events, and all cause death in people with no known cardiovascular disease or at high risk of cardiovascular disease for the purpose of primary prevention. We also assessed the modifying effects of fish oil supplements on the disease process, from atrial fibrillation to other outcomes, in people with known cardiovascular disease for the purpose of secondary prevention.

The UK Biobank is a community based cohort study with more than half a million UK inhabitants aged 40-69 years at recruitment. 11–13 Participants were invited to participate in this study if they were registered with the NHS and lived within 35 km of one of 22 Biobank assessment centres. Between 1 March 2006 and 31 July 2010, a baseline survey was conducted, based on a touch screen questionnaire and face-to-face interviews, to collect detailed personal, socioeconomic, and lifestyle characteristics, and information on diseases. 11–13

We excluded patients who had a diagnosis of atrial fibrillation (n=8326), heart failure (n=2748), myocardial infarction (n=11 949), stroke (n=7943), or cancer (n=48 624) at baseline; who withdrew from the study during follow-up (n=1299); or who had incomplete or outlier data for the main information (n=11 748). Because we focused only on a specific sequence of progression of cardiovascular disease (ie, from healthy status to atrial fibrillation, to major adverse cardiovascular events, and then to death), we excluded 1983 participants with other transition patterns. The remaining 415 737 participants were included in this analysis ( figure 1 ).

• Download figure
• Open in new tab
• Download powerpoint

Flowchart of selection of participants in study. The count of diagnosed diseases does not equate to the total number of individuals, because each person could have multiple diagnoses

Determining use of fish oil supplements

Information on regular use of fish oil supplements was collected from a self-reported touchscreen questionnaire during the baseline survey. 14 15 Each participant was asked whether they regularly used any fish oil supplement. Trained staff conducted a verbal interview with participants, asking if they were currently receiving treatments or taking any medicines, including omega 3 or fish oil supplements. Based on this information, we classified participants as regular users of fish oil supplements and non-users.

Follow-up and outcomes

Participants were followed up from the time of recruitment to death, loss to follow-up, or the end date of follow-up (31 March 2021), whichever came first. Incident cases of interest, including atrial fibrillation, heart failure, stroke, and myocardial infarction, were identified by linkage to death registries, primary care records, and hospital inpatient records. 11 Information on deaths was obtained from death registries of the NHS Information Centre, for participants in England and Wales, and from the NHS Central Register Scotland, for participants in Scotland. 11 Outcomes were defined by a three character ICD-10 (international classification of diseases, 10th revision) code. In this study, atrial fibrillation was defined by ICD-10 code I48, and major adverse cardiovascular events was determined by a combination of heart failure (I50, I11.0, I13.0, and I13.2), stroke (I60-I64), and myocardial infarction (I21, I22, I23, I24.1, and I25.2) codes.

We collected baseline data on age (<65 years and ≥65 years), sex (men and women), ethnic group (white and non-white), Townsend deprivation index (with a higher score indicating higher levels of deprivation), smoking status (never, previous, and current smokers), and alcohol consumption (never, previous, and current drinkers). Data for sex were taken from information in UK Biobank rather than from patient reported gender. Baseline dietary data were obtained from a dietary questionnaire completed by the patient or by an interviewer. The questionnaire was established for each nation (ie, England, Scotland, and Wales) to assess an individual's usual food intake (oily fish, non-oily fish, vegetables, fruit, and red meat). Diabetes mellitus was defined by ICD-10 codes E10-E14, self-reported physician's diagnosis, self-reported use of antidiabetic drugs, or haemoglobin A1c level ≥6.5% at baseline. Hypertension was defined by ICD-10 code I10 or I15, self-reported physician's diagnosis, self-reported use of antihypertensive drugs, or measured systolic and diastolic blood pressure ≥130/85 mm Hg at baseline. Information on other comorbidities (obesity (ICD-10 code E66), chronic obstructive pulmonary disease (J44), and chronic renal failure (N18)) was extracted from the first occurrence (UKB category ID 1712). Information on the use of drugs, including antihypertensive drugs, antidiabetic drug, and statins, was extracted from treatment and drug use records. Biochemistry markers were measured immediately at the central laboratory from serum samples collected at baseline. Binge drinking was defined as consumption of ≥6 standard drinks/day for women or ≥8 standard drinks/day for men. Detailed information on alcohol consumption and binge drinking in the UK Biobank was reported previously. 16

Statistical analysis

Characteristics of participants are summarised as number (percentages) for categorical variables and mean (standard deviation (SD)) for continuous variables. Comparisons between regular users of fish oil supplements and non-users were made with the χ 2 test or Student's t test.

We used a multi-state regression model to assess the role of regular use of fish oil supplements in the temporal disease progression from healthy status to atrial fibrillation, to major adverse cardiovascular events, and subsequently to death. The multi-state model is an extension of competing risks survival analysis. 17–19 The model allows simultaneous estimation of the role of risk factors in transitions from a healthy state to atrial fibrillation (transition A), healthy state to major adverse cardiovascular events (transition B), healthy state to death (transition C), atrial fibrillation to major adverse cardiovascular events (transition D), atrial fibrillation to death (transition E), and major adverse cardiovascular events to death (transition F) (transition pattern I, figure 2 ). The focus on these six transitions rather than on all possible health state transitions was preplanned and evidence based. If participants entered different states on the same date, we used the date of the theoretically previous state as the entry date of the latter state minus 0.5 days.

Numbers of participants in transition pattern I, from baseline to atrial fibrillation, major adverse cardiovascular events, and death

We further examined the effects of regular use of fish oil supplements on other pathways. For example, we divided major adverse cardiovascular events into three individual diseases (heart failure, stroke, and myocardial infarction), resulting in three independent pathways (transition patterns II, III, and IV, online supplemental figures S1–S3 ). All models were adjusted for age, sex, ethnic group, Townsend deprivation index, consumption of oily fish, consumption of non-oily fish, smoking status, alcohol consumption, obesity, hypertension, diabetes mellitus, chronic obstructive pulmonary disease, chronic renal failure, and use of statins, antidiabetic drugs, and antihypertensive drugs.

Supplemental material

We conducted several sensitivity analyses for the multi-state analyses of transition pattern A: additionally adjusting for setting (urban and rural), body mass index (underweight, normal, overweight, and obese), and physical activity (low, moderate, and high) in the model; adjusting for binge drinking rather than alcohol consumption; additionally adjusting for other variables of dietary intake (consumption of vegetables, fruit, and red meat); calculating participants' entry date into the previous state with different time intervals (0.5 years, one year, and two years); excluding participants who entered different states on the same date; excluding events occurring in the first two years of follow-up; restricting the follow-up date to 31 March 2020 to evaluate the influence of the covid-19 pandemic; and the use of the inverse probability weighted method to deal with biases between the regular users and non-users of fish oil supplements. Also, we conducted grouped analyses for sex, age group, ethnic group, smoking status, consumption of oily fish, consumption of non-oily fish, hypertension, and drug use, to examine effect modification. The interactions were tested with the likelihood ratio test. All analyses were carried out with R software (version 4.0.3), and the multi-model analysis was performed with the mstate package. A two tailed P value <0.05 was considered significant.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Participants were involved in developing the ethics and governance framework for UK Biobank and have been engaged in the progress of UK Biobank through follow-up questionnaires and additional assessment visits. UK Biobank keeps participants informed of all research output through the study website ( https://www.ukbiobank.ac.uk/explore-your-participation ), participant events, and newsletters.

A total of 415 737 participants (mean age 55.9 (SD 8.1) years; 55% women), aged 40-69 years, were analysed, and 31.4% (n=1 30 365) of participants reported regular use of fish oil supplements at baseline ( figure 1 ). Table 1 shows the characteristics of regular users (n=130 365) and non-users (n=285 372) of fish oil supplements. In the group of regular users of fish oil supplements, we found higher proportions of elderly people (22.6% v 13.9%), white people (95.1% v 94.2%), and women (57.6% v 53.9%), and higher consumption of alcohol (93.1% v 92.0%), oily fish (22.1% v 15.4%), and non-oily fish (18.0% v 15.4%) than non-users. The Townsend deprivation index (mean −1.5 (SD 3.0) v −1.3 (3.0)) and the proportion of current smokers (8.1% v 11.4%) were lower in regular users of fish oil supplements. Online supplemental table S1 provides more details on patient characteristics and online supplemental table S2 compares the basic characteristics of included and excluded people.

• View inline

Baseline characteristics of study participants grouped by use of fish oil supplements

Over a median follow-up time of of 11.9 years, 18 367 participants had atrial fibrillation (transition A) and 17 826 participants had major adverse cardiovascular events (transition B); 14 902 participants died without having atrial fibrillation or major adverse cardiovascular events (transition C). Among patients with incident atrial fibrillation, 4810 developed major adverse cardiovascular events (transition D) and 1653 died (transition E). Among patients with incident major adverse cardiovascular events, 5585 died during follow-up (transition F, figure 2 ). In separate analyses for individual diseases (transition patterns II, III, and IV, online supplemental figures S1–S3 ), in patients with atrial fibrillation, 3085 developed heart failure, 1180 had a stroke, and 1415 had a myocardial infarction. During follow-up, 2436, 2088, and 2098 deaths occurred in patients with heart failure, stroke, and myocardial infarction, respectively.

Multi-state regression results

Table 2 shows the different roles of regular use of fish oil supplements in transitions from healthy status to atrial fibrillation, to major adverse cardiovascular events, and then to death. For individuals in the primary stage (healthy status), we found that the use of fish oil supplements had a harmful effect on the transition from health to atrial fibrillation, with an adjusted hazard ratio of 1.13 (95% CI 1.10 to 1.17, transition A). The hazard ratio for transition B (from health to major adverse cardiovascular events) was 1.00 (95% CI 0.97 to 1.04) and for transition C (from health to death) was 0.98 (0.95 to 1.02).

Hazard ratios (95% confidence intervals) for each transition, for different transition patterns for progressive cardiovascular disease by regular use of fish oil supplements

For individuals in the secondary stage (atrial fibrillation) at the beginning of the study, regular use of fish oil supplements decreased the risk of major adverse cardiovascular events (transition D, hazard ratio 0.92, 95% CI 0.87 to 0.98), and had a borderline protective effect on the transition from atrial fibrillation to death (transition E, 0.91, 0.82 to 1.01). For transition F, from major adverse cardiovascular events to death, after adjusting for covariates, the hazard ratio was 0.99 (0.94 to 1.06, transition pattern I, table 2 ).

We divided major adverse cardiovascular events into three individual diseases (ie, heart failure, stroke, and myocardial infarction) and found that regular use of fish oil supplements was marginally associated with an increased risk of stroke in people with a healthy cardiovascular state (hazard ratio 1.05, 95% CI 1.00 to 1.11), whereas a protective effect was found in transitions from healthy cardiovascular states to heart failure (0.92, 0.86 to 0.98). For patients with atrial fibrillation, we found that the beneficial effects of regular use of fish oil supplements were for transitions from atrial fibrillation to myocardial infarction (0.85, 0.76 to 0.96), and from atrial fibrillation to death (0.88, 0.81 to 0.95) for transition pattern IV. For patients with heart failure, we found a protective effect of regular use of fish oil supplements on the risk of mortality (0.91, 0.84 to 0.99) (transition patterns II, III, and IV, table 2 ).

Stratified and sensitivity analyses

We found that age, sex, smoking, consumption of non-oily fish, prevalent hypertension, and use of statins and antihypertensive drugs modified the associations between regular use of fish oil supplements and the transition from healthy states to atrial fibrillation ( online supplemental figure S4 ). We found that the association between regular use of fish oil supplements and risk of transition from healthy states to major adverse cardiovascular events was greater in women (hazard ratio 1.06, 95% CI 1.00 to 1.11, P value for interaction=0.005) and non-smoking participants (1.06, 1.06 to 1.11, P value for interaction=0.001) ( online supplemental figure S4 ). The protective effect of regular use of fish oil supplements on the transition from healthy states to death was greater in men (hazard ratio 0.93, 95% CI 0.89 to 0.98, P value for interaction=0.003) and older participants (0.91, 0.86 to o 0.96, P value for interaction=0.002) ( online supplemental figures S5 and S6 ). The results were not substantially changed in the sensitivity analyses ( online supplemental table S3 ).

Principal findings

Our study characterised the regular use of fish oil supplements on the progressive course of cardiovascular disease, from a healthy state (primary stage), to atrial fibrillation (secondary stage), major adverse cardiovascular events (tertiary stage), and death (end stage). In this prospective analysis of more than 400 000 UK adults, we found that regular use of fish oil supplements could have a differential role in the progression of cardiovascular disease. For people with a healthy cardiovascular profile, regular use of fish oil supplements, a choice of primary prevention, was associated with an increased risk of atrial fibrillation. For participants with a diagnosis of atrial fibrillation, however, regular use of fish oil supplements, as secondary prevention, had a protective effect or no effect on transitions from atrial fibrillation to major adverse cardiovascular events, atrial fibrillation to death, and major adverse cardiovascular events to death. When we divided major adverse cardiovascular events into three individual diseases (ie, heart failure, stroke, and myocardial infarction), we found associations that could suggest a mildly harmful effect between regular use of fish oil supplements and transitions from a healthy cardiovascular state to stroke, whereas potential beneficial associations were found between regular use of fish oil supplements and transitions from atrial fibrillation to myocardial infarction, atrial fibrillation to death, and heart failure to death.

Comparison with other studies

Primary prevention.

The cardiovascular benefits of regular use of fish oil supplements have been examined in numerous studies but the results are controversial. Extending previous reports, our study estimated the associations between regular use of fish oil supplements and specific clinical cardiovascular disease outcomes in people with no known cardiovascular disease. Our findings are in agreement with the results of several previous randomised controlled trials and meta-analyses. The Long-Term Outcomes Study to Assess Statin Residual Risk with Epanova in High Cardiovascular Risk Patients with Hypertriglyceridaemia (STRENGTH) reported that consumption of 4 g/day of marine omega 3 fatty acids was associated with a 69% higher risk of new onset atrial fibrillation in people at high risk of cardiovascular disease. 20 A meta-analysis of seven randomised controlled trials showed that users of marine omega 3 fatty acids supplements had a higher risk of atrial fibrillation events, with a hazard ratio of 1.25 (95% CI 1.07 to 1.46, P=0.013). 21 The Vitamin D and Omega-3 Trial (VITAL Rhythm study), a large trial of omega 3 fatty acids for the primary prevention of cardiovascular disease in adults aged ≥50 years, however, found no effects on incident atrial fibrillation, major adverse cardiovascular events, or cardiovascular disease mortality among those treated with 840 mg/day of marine omega 3 fatty acids compared with placebo. 10 22

One possible explanation for the inconsistent results in these studies is that adverse effects might be related to dose and composition. Higher doses of omega 3 fatty acids used in previous studies might have had an important role in causing an adverse effect on atrial fibrillation. 21 One study found that high concentrations of fish oil altered cell membrane properties and inhibited Na-K-ATPase pump activity, whereas a low concentration of fish oil minimised peroxidation potential and optimised activity. 23 In another study, individuals with atrial fibrillation or flutter had higher percentages of total polyunsaturated fatty acids, and n-3 and n-6 polyunsaturated fatty acids, on red blood cell membranes than healthy controls. 24

In terms of composition of omega 3 fatty acids, a recent meta-analysis showed that eicosapentaenoic acid alone can be more effective at reducing the risk of cardiovascular disease than the combined effect of eicosapentaenoic acid and docosahexaenoic acid. 25 Similar outcomes were reported in the INSPIRE study, which showed that higher levels of docosahexaenoic acid reduced the cardiovascular benefits of eicosapentaenoic acid when given as a combination. 26 Another possible explanation is that age, sex, ethnic group, smoking status, dietary patterns, and use of statins and antidiabetic drugs by participants might modify the effects of regular use of fish oil supplements on cardiovascular disease events. Despite these differences in risk estimates, our findings do not support the use of fish oil or omega 3 fatty acid supplements for the primary prevention of incident atrial fibrillation or other specific clinical cardiovascular disease events in generally healthy individuals. Caution might be warranted when fish oil supplements are used for primary prevention because of the uncertain cardiovascular benefits.

Secondary prevention

Our large scale cohort study assessed the role of regular use of fish oil supplements on the disease process, from atrial fibrillation to more serious cardiovascular disease stages, to death, in people with known cardiovascular disease. Contrary to the observations for primary prevention, we found associations that could suggest beneficial effects between regular use of fish oil supplements and most cardiovascular disease transitions. No associations were found between regular use of fish oil supplements and transitions from atrial fibrillation to death, or from major adverse cardiovascular events to death.

Consistent with our hypothesis, the Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico (GISSI) Prevenzione study reported an association between administration of low dose prescriptions of n-3 polyunsaturated fatty acids and reduced cardiovascular events in patients with recent myocardial infarction. 27 A meta-analysis of 16 randomised controlled trials also reported a tendency towards a greater beneficial effect for secondary prevention in patients with cardiovascular disease. 28 Why patients with previous atrial fibrillation benefit is unclear. These findings indicate that triglyceride independent effects of omega 3 fatty acids might in part be responsible for the benefits in cardiovascular disease seen in previous trials. 29–31 No proven biological mechanism for this explanation exists, however, and the dose and formulation of omega 3 fatty acids used in clinical practice are not known.

For the disease process, from cardiovascular disease to death, our findings are consistent with the results of secondary prevention trials of omega 3 fatty acids, which have mostly shown a weak or neutral preventive effect in all cause mortality with oil fish supplements. The GISSI heart failure trial (GISSI-HF), conducted in 6975 patients with chronic heart failure, reported that supplemental omega 3 fatty acids reduced the risk of all cause mortality by 9% (hazard ratio 0.91, 95% CI 0.833 to 0.998, P=0.041). 32 Zelniker et al showed that omega 3 fatty acids were inversely associated with a lower incidence of sudden cardiac death in patients with non-ST segment elevation acute coronary syndrome. 33 A meta-analysis found that use of omega 3 supplements of ≤1 capsule/day was not associated with all cause mortality, but among participants with a risk of cardiovascular disease, taking a higher dose was associated with a reduction in cardiac death and sudden death. 28 Individuals who might benefit the most from fish oil or omega 3 fatty acid supplements are possibly more vulnerable individuals, such as those with previous cardiovascular diseases and those who can no longer live in the community. How fish oil supplements stop further deterioration of cardiovascular disease is unclear, but the theory that supplemental omega 3 fatty acids might protect the coronary artery is biologically plausible, suggesting that omega 3 fatty acids have anti-inflammatory and anti-hypertriglyceridaemia effects, contributing to a reduction in thrombosis and improvement in endothelial function. 34–41 Nevertheless, the effects of omega 3 fatty acids vary according to an individual's previous use of statins, which might partly explain the different effects of fish oil supplements in people with and without cardiovascular disease.

Many studies of omega 3 fatty acids, including large scale clinical trials and meta-analyses, have not produced entirely consistent results. 21 25 42 Our study mainly explored the varied potential effects of regular use of fish oil supplements on progression of cardiovascular disease, offering an initial overview of this ongoing discussion. Our findings suggest caution in the use of fish oil supplements for primary prevention because of the uncertain cardiovascular benefits and adverse effects. Further studies are needed to determine whether potential confounders modify the effects of oil fish supplements and the precise mechanisms related to the development and prognosis of cardiovascular disease events.

Strengths and limitations of this study

The strengths of our study were the large sample size, long follow-up period, which allowed us to analyse clinically diagnosed incident diseases, and complete data on health outcomes. Another strength was our analytical strategy. The multi-state model gives less biased estimates than the conventional Cox model, and distinguished the effect of regular use of fish oil supplements on each transition in the course of cardiovascular disease.

Our study had some limitations. Firstly, as an observational study, no causal relations can be drawn from our findings. Secondly, although we adjusted for multiple covariates, residual confounding could still exist. Thirdly, information on dose and formulation of the fish oil supplements was not available in this study, so we could not evaluate potential dose dependent effects or differentiate between the effects of different fish oil formulations. Fourthly, the use of hospital inpatient data for determining atrial fibrillation events could have excluded some events triggered by acute episodes, such as surgery, trauma, and similar conditions, resulting in underestimation of the true risk because undiagnosed atrial fibrillation is a common occurrence. 43 Fifthly, most of the participants in this study were from the white ethnic group and whether the findings can be generalised to other ethnic groups is not known. Finally, our study did not consider behavioural changes in populations with different cardiovascular profiles because of limited information, and variations in outcomes for different cardiovascular states merits further exploration.

Conclusions

This large scale prospective study of a UK cohort suggested that regular use of fish oil supplements might have differential roles in the course of cardiovascular diseases. Regular use of fish oil supplements might be a risk factor for atrial fibrillation and stroke among the general population but could be beneficial for disease progression, from atrial fibrillation to major adverse cardiovascular events, and from atrial fibrillation to death. Further studies are needed to determine whether potential confounders modify the effects of oil fish supplements and the precise mechanisms for the development and prognosis of cardiovascular disease events.

Ethics statements

Patient consent for publication.

Consent obtained directly from patients.

Ethics approval

The UK Biobank study obtained ethical approval from the North West Multicentre Research ethics committee, Information Advisory Group, and the Community Health Index Advisory Group (REC reference for UK Biobank 11/NW/0382). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

This study was conducted with UK Biobank Resource (application No: 69550). We appreciate all participants and professionals contributing to UK Biobank.

• Mensah GA ,
• Johnson CO , et al
• Gao MM , et al
• Saravanan P ,
• Davidson NC ,
• Schmidt EB , et al
• National Institute for Health and Care Excellence
• Lichtenstein AH ,
• Vadiveloo M , et al
• Djuricic I ,
• Miller M , et al
• Kesse-Guyot E ,
• Czernichow S , et al
• Klemsdal TO ,
• Sandvik L , et al
• Manson JE ,
• Lee I-M , et al
• Gallacher J ,
• Allen N , et al
• Littlejohns TJ ,
• Sudlow C , et al
• Allen NE , et al
• Zhong W-F ,
• Liu S , et al
• Wu Z , et al
• Gallagher C ,
• Elliott AD , et al
• Qian SE , et al
• Nicholls SJ ,
• Lincoff AM ,
• Garcia M , et al
• Djousse L ,
• Al-Ramady OT , et al
• Bassuk SS ,
• Cook NR , et al
• Cazzola R ,
• Della Porta M ,
• Castiglioni S , et al
• Viviani Anselmi C ,
• Ferreri C ,
• Novelli V , et al
• Khan MS , et al
• Knowlton K , et al
• Marchioli R ,
• Bomba E , et al
• Olmastroni E ,
• Gazzotti M , et al
• Al Rifai M , et al
• Tavazzi L ,
• Maggioni AP ,
• Marchioli R , et al
• Zelniker TA ,
• Morrow DA ,
• Scirica BM , et al
• Limonte CP ,
• Zelnick LR ,
• Ruzinski J , et al
• Nelson JR ,
• Miller PE ,
• Van Elswyk M ,
• Alexander DD
• Mozaffarian D ,
• Bornfeldt KE
• Harris WS ,
• Ginsberg HN ,
• Arunakul N , et al
• Markozannes G ,
• Tsapas A , et al
• Svennberg E ,
• Engdahl J ,
• Al-Khalili F , et al

Supplementary materials

Supplementary data.

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

• Data supplement 1
• Data supplement 2

GYL and HL are joint senior authors.

Contributors HL supervised the whole project and designed the work. GC and HL directly accessed and verified the underlying data reported in the manuscript. GC contributed to data interpretation and writing of the report. ZQ, SZ, JZ, ZZ, MGV, HEA, CW, and GYHL contributed to the discussion and data interpretation, and revised the manuscript. All authors had full access to all of the data in the study and had final responsibility for the decision to submit for publication. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. HL is the guarantor. Transparency: The lead author (guarantor) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Funding This work was supported by the Bill and Melinda Gates Foundation (grant No INV-016826). Under the grant conditions of the foundation, a creative commons attribution 4.0 generic license has already been assigned to the author accepted manuscript version that might arise from this submission. The funder had no role in considering the study design or in the collection, analysis, interpretation of data, writing of the report, or decision to submit the article for publication.

Competing interests All authors have completed the ICMJE uniform disclosure form at www.icmje.org/disclosure-of-interest/ and declare: support from Bill and Melinda Gates Foundation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Read the full text or download the PDF: